Skip to main content

Full text of "Davidson/Eternity"

See other formats


( 



( 
( 
( 

k 
|( 

!( 

Dkitiiizido pin uso 
p*jr<nl. Iivor nio 
disfribuir* 



( 



Proofs for Eternity, Creation 
and the Existence of God 
in Medieval Islamic 
and Jewish Philosophy y v ^ 



HERBERT A. DAVIDSON 
University of California, Los Angeles 



C 



New York * Oxford 
OXFORD UNIVERSITY PRESS 
1987 



The von Grunebaum Center 
for Near Eastern Studies 



PROOFS FOR ETERNITY, CREATION AND 
THE EXISTENCE OF GOD IN MEDIEVAL ISLAMIC 
AND JEWISH PHILOSOPHY 



Oxford University Press 



Oxford New York Toronto 
Delhi Bomby Calcutta Madras Karachi 
Petaling Jaya ; Singapore Hong Kong Tokyo 
Nairobi Dar es Salaam Cape Town 
Melbourne Auckland 
and associated companies in 
Beirut Berlin Ibadan Nicosia 

Copyright © 1987 by Herbert A. Davidson 

Published by Oxford University Press, Inc. , 
200 Madison Avenue, New York, New York 10016 

Oxford is a registered trademark of Oxford University Press 

Library of Congress Cataloging-in-Publication Data 
Davidson, Herbert A. (Herbert Alan) 
Proofs for eternity, creation, and the existence 
of God in medieval Islamic and Jewish philosophy. 
New York: Oxford University Press, 1987. 
Includes indexes. 
1. God (Islam) — Proof— History of doctrines. 
2. God (Judaism) — History of doctrines. 3. Philosophy, Islamic. 
4. Philosophy, Jewish. I. Title. 
B745.G63D38 1987 212'.1'0902 86-33179 
ISBN 0-19-504953-5 

Published with the assistance of The Louis and Minna Epstein Fund of 
the American Academy for Jewish Research 



13579 10 8642 

Printed in the United States of America 
on acid-free paper 



For 

Rachel and Jessica 



Acknowledgments 



Chapters IV and V of the present book are a reworking of an article that appeared 
in the Journal of the American Oriental Society, Volume 89. Chapter VI incor- 
porates material from articles appearing in Philosophy East and West, Volume 18, 
and Studies in Medieval Jewish History and Literature (Cambridge, 1979). An 
expanded version of Appendix A, part 2, appeared in Studies in Jewish Religious 
and Intellectual History Presented to Alexander Altmann (University, Alabama, 
1979). 

The National Endowment for the Humanities and the UCLA Academic Senate 
have supported my work; the von Grunebaum Center for Near Eastern Studies, 
UCLA, took responsibility for preparing a camera-ready copy; a grant from the 
American Academy for Jewish Research to the von Grunebaum Center helped to 
defray some of the composition costs. I wish to express my warmest thanks to all 
those institutions. I also wish to thank Marina Preussner of the von Grunebaum 
Center for her invaluable aid and irrepressible good cheer. 

My wife subjected the book to a painstaking, and often painful, critique. Any 
clarity that is to be detected is her doing. 

The book was complete in 1980. 



( 

f 
I 
( 
( 
( 
( 
( 
f 

r 
( 
( 
( 
( 
( 
( 
( 
( 
( 
< 
: 
r 

( 
( 
( 
( 
( 
( 

( 
( 
f 
( 



Contents 



I Introduction 1 

1. Eternity, creation, and the existence of God 1 

2. The present book 6 

II Proofs of Eternity from the Nature of the World 9 

1. Proofs of eternity 9 

2. Proofs of eternity from the nature of the physical world 12 

3. Replies to proofs from the nature of the world 30 

4. Summary 4° 

III Proofs of Eternity from the Nature of God 49 

1. The proofs 4 9 

2. Replies to proofs from the nature of the cause of the 
universe 67 

3. Summary 85 

IV John Philoponus' Proofs of Creation and Their Entry into 
Medieval Arabic Philosophy 86 

1. Philoponus' proofs of creation 86 

2. Saadia and Philoponus 95 

3. Kindi and Philoponus 106 

4. Summary 



x Contents 

V Kalam Proofs for Creation 117 

1. Proofs from the impossibility of an infinite number 117 

2. Responses of the medieval Aristotelians to proofs of 
creation from the impossibility of an infinite number 127 

3. The standard Kalam proof for creation: the proof from 
accidents 134 

4. Juwayni's version of the proof from accidents 143 

5. Proofs from composition 146 

{ VI Arguments from the Concept of Particularization 154 

1. Inferring the existence of God from creation 154 

2. Arguments from the concept of particularization 174 

3. Particularization arguments for the existence of God 
without the premise of creation; particularization 
arguments for creation 187 

4. Ghazali and Maimonides 194 

5. Additional arguments for creation in Maimonides and 
Gersonides 203 

VII Arguments from Design 213 

1. Cosmological, ideological, and ontological proofs of the 
existence of God 213 

2. Tcleological arguments 216 

3. Summary 236 

VIII The Proof from Motion 237 

1. Aristotle's proof from motion 237 

2. Maimonides' version of the proof from motion 240 

3. Hasdai Crescas' critique of the proof from motion 249 

4. Another proof from motion 275 



IX Avicenna's Proof of the Existence of a Being Necessarily 

Existent by Virtue of Itself 281 



Contents xi 

1. First cause of motion and first cause of existence 281 

2. The existence of God: a problem for metaphysics 284 

3. Necessarily existent being and possibly existent being 289 

4. The attributes of the necessarily existent by virtue of itself 293 

5. Proof of the existence of the necessarily existent by virtue 

of itself 298 

6. Questions raised by Avicenna's proof 304 

7. The version of Avicenna's proof in ShahrastanI and 

Crescas 307 

8. Summary 309 

X Averroes' Critique of Avicenna's Proof 311 

1. The proof of the existence of God as a subject for 

physics 312 

2. Necessarily existent by virtue of another, possibly 

existent by virtue of itself 318 

3. The nature of the celestial spheres according to Averroes 321 

4. Averroes' critique of the body of Avicenna's proof 331 

5. Summary 334 

XI Proofs of the Existence of God from the Impossibility of an 
Infinite Regress of Efficient Causes 336 

1. The proof from the impossibility of an infinite regress of 
causes 336 

2. Unity and incorporcality 345 

3. The proof from the impossibility of an infinite regress of 
efficient causes and the proof from the concepts possibly 
existent and necessarily existent 350 

4. Resume 362 

5. Crescas on the impossibility of an infinite regress 365 

6. Ghazali's critique of Avicenna's proof 366 

7. Summary 375 



xii 



Contents 



XII Subsequent History of Proofs from the Concept of Necessary 

Existence 378 

1. Maimonidcs and Aquinas 378 

2. The influence of Avicenna's proof 385 

3. Proofs of the existence of God as a necessarily existent 

being in modern European philosophy 388 

4. Summary 405 

5. Concluding remark 406 

Appendix A. Two Philosophic Principles 407 

1. The principle that an infinite number is impossible 407 

2. The principle that a finite body contains only finite power 409 

Appendix B. Inventory of Proofs 412 

Primary Sources 414 

Index of Philosophers 421 

Index of Terms 427 



PROOFS FOR ETERNITY, CREATION AND 
THE EXISTENCE OF GOD IN MEDIEVAL ISLAMIC 
AND JEWISH PHILOSOPHY 



r 



i 
( 

' ( 
( 
( 
( 
( 
( 

( 
i 
i 
i 
( 
( 
< 
( 
( 
i 

( 
i 

\{ 
\' 
i 

;/ 

\( 
I 

/ 
( 
f 



I 

Introduction 



1. Eternity, creation, and the existence of God 

One might well expect the existence of God to be the initial issue of natural 
theology; but such was not the case in the Islamic and Jewish Middle Ages. 
Medieval Muslim and Jewish philosophers did as a matter of course construct 
their natural theology and their metaphysics in general on the existence of God. 
The provability of the deity's existence was, furthermore, disputed; for whereas 
most philosophers were confident that the deity's existence can be demonstrated 
rationally, some demurred. When the possibility of demonstrating the existence 
of God was challenged, the challenge came, however, not from radicals who 
doubted the proposition, but from conservatives who questioned the competence 
of human reason to demonstrate it. While the provability of God's existence 
might, then, be subject to dispute, God's existence never was, and the Middle 
Ages were free from atheism and agnosticism, at least public atheism and agnos- 
ticism, on the philosophic plane. The existence of God could not, as a conse- 
quence, be the initial issue for natural theology. The existence of God, as distinct 
from the provability of God's existence, was not strictly an issue at all. 

The initial issue of natural theology for Muslims and Jews, the most funda- 
mental issue where opinions divided, was, it may be ventured, the inquiry con- 
cerning whether the world is eternal or had a beginning. Much more is at stake 
there than chronology or hermeneutics — the age of the universe or the question 
whether the scriptural account of the genesis of the universe should be taken 
literally or allegorically. The issue of eternity and creation 1 provided an arena 
for determining the relationship of God to the universe, for determining, specif- 
ically, whether God is a necessary or a voluntary cause. If the world should be 
eternal, and a deity is recognized, the deity's relationship to the universe would 
likewise be eternal. Since eternity and necessity are, by virtue of an Aristotelian 

1 1 employ the term creation to mean (he thesis that the world came into existenee after not 
having existed, not the more specific thesis that a creator brought the world into existence. Medieval 
thinkers who accepted the former thesis were invariably certain that the latter thesis can be inferred 
from it. 



1 



2 



Introduction 



rule, mutually implicative, 2 an eternal relationship is a relationship bound by 
necessity; and necessity excludes wjll. 3 The eternity of the world thus would 
imply that the deity is, as the cause of the universe, 4 bereft of will. A beginning 
of the world would, by contrast, lead to a deity possessed of will. Should the 
world be understood to have a beginning, all medieval thinkers agreed, the exis- 
tence of a creator can be inferred; and the decision on the creator's part to bring 
a world into existence where no world existed before would constitute a supreme 
and paradigmatic act of volition. Will in the deity would, therefore, be ruled out 
by the eternity of the world and entailed by creation. 

The issue of eternity and creation often intertwined with the enterprises of 
proving the existence of God. The majority of Islamic and Jewish proofs for the 
existence of God take either eternity or creation as a premise, and require a 
resolution of that issue before their own proper subject can be broached. In the 
Aristotelian proof of the existence of God as the prime mover and in kindred 
proofs, the eternity of the world is an indispensable premise and must be estab- 
lished prior to the proof itself. The world is shown to be eternal, and the eternal 
motion or the eternal existence of the universe is shown to have a cause, which 
is identified as the deity. 5 In Kalam proofs of the existence of God, the indis- 
pensable premise is creation. The Kalam thinkers followed what has been called 
the Platonic procedure, 6 that is, the procedure of first proving the creation of the 
world and then inferring therefrom the existence of a creator, again identified as 
the deity. 7 

The decision to demonstrate the existence of God from the premise of eternity, 
on the one hand, or from the premise of creation, on the other, was not the result 
merely of one party's happening to be convinced of the truth of eternity and the 
other's being convinced of the truth of creation. The decision was connected with 
the diverse conceptions of the deity which accompany the two premises. The 
choice of one or the other premise would be reflected in the conclusion of a proof 

2 Aristotlc, De Generalione II, II, 338a, 1. 

3 Plotinus speaks of a "necessary free will" (Enneads, IV, 8, 5); and Aviccnna and Avcrrocs 
speak of the deity's "eternal will" (Avicenna, Shifa' : Ilahiyat, ed. G. Anawati and S. Zayed [Cairo, 
1960], p. 366; Averroes, K. al-Kashf, ed. M. Mueller [Munich, 1859], p. 52; German translation, 
with pagination of the Arabic indicated: Philosophic und Theologie von Averroes, trans. M. Mueller 
[Munich, 1875]). Crescas, Or ha-Shem, III, i, 5, also defends the possibility of an eternal will in 
God. But these are Pickwickian senses of will; cf. Maimonides, Guide to the Perplexed, II, 21. 

4 ConceivabIy the deity could, though the universe is eternal, exercise his will in some fashion 
that docs not relate to the universe. But that thought, if it has any meaning at all, was not entertained 
by the medicvals. 

3 Cf. below, p. 239. 

6 Cf. Plato, Timaeus, 28; Moses Narboni, Commentary on Maimonides' Guide (Vienna, 1853). 
II, 2; C. Baeumker, Witelo (Muenster, 1908), pp. 320 ff.; H. Wolfson, "Notes on Proofs of the 
Existence of God in Jewish Philosophy," reprinted in his Studies in the History of Philosophy and 
Religion, Vol. I (Cambridge, Mass., 1973), pp. 571-572. 

7 Cf. below, pp. 154 ff. 



Introduction 



3 



for the existence of God, since, as already seen, a proof from the premise of 
eternity would lead to a deity bound by necessity and a proof from the premise 
of creation would lead to a deity possessed of will. Something more might be 
involved. Proofs proceeding from the two different premises can differ in their 
understanding of what constitutes a genuine proof of the existence of God. Every 
proof of the existence of God must at some stage, whether explicitly or—as 
occurs far more often — implicitly, presuppose a definition of God, a set of spec- 
ifications requisite and sufficient for the deity. A proof of the existence of God 
is a chain of reasoning which concludes with the existence of a being distinguished 
by certain attributes. Unless the attributes qualify the being possessing them as 
the deity, no proof, however correctly reasoned, can claim that the being whose 
existence it arrives at is God. Proofs of the existence of God from the premise 
of eternity, at least among Islamic and Jewish philosophers, implicitly assume or 
explicitly state three specifications for the deity. By deity, a being is meant which 
is, firstly, an uncaused cause; secondly, incorporeal; and thirdly, one. 8 Any chain 
of reasoning concluding with the existence of a single, uncaused, incorporeal 
cause would accordingly constitute a successful proof of the existence of God. 
Volition is not, in proofs from the premise of eternity, included among the spec- 
ifications. In fact, volition is ruled out, since, as has been seen, the deity would 
be bound by necessity if the world is eternal. As Ghazali explains the virtue of 
the proof from the opposite premise, the premise of creation, that proof must be 
resorted to because of the inadequacy of the definition just given. To be an 
uncaused cause, incorporeal, and one, Ghazali insists, is a good deal less than 
to be the deity. For nothing could conceivably be designated as the deity if it is 
unable to make decisions affecting the course of events in the universe. 9 The 
specifications presupposed in proofs from the premise of eternity must, on this 
view, be supplemented with a further attribute, volition. 

Ghazali had an additional reason for rejecting any proof of the existence of 
God not based on the premise of creation. He contends that the notion of eternal 
causation is intrinsically nonsensical, that what comes into existence after not 
existing can alone be thought of as having a cause. To advocate eternity would 
thus be tantamount to denying a cause of the universe, and to countenance eternity 
would be tantamount to countenancing the causelessness of the universe. Argu- 
mentation that does not employ the premise of creation hence would fail to prove 
the existence of God not merely for those who number volition among the req- 
uisite specifications. It would fail to prove the existence of God even on the view 
of the proponents of the proofs from eternity; for it would fail to establish the 

"implicitly in Maimonides. Guide, I, 71. Also cf. Alfarabi. A'. Ara AM al-Madina al-radila, 
ed. F. Dictcrici (Leiden, 1895), beginning; German translation: Der Musterstaat, trans. F. Dietcrici 
(Leiden, 1900). Aviccnna, Shifa' : Ilahiyat, pp. 37-47. 

"Ghazali. Tahafut al-liihisifa, cd. M. Houyges (Beirut, 1927). III. §§3, 16; English translation 
in Averroes' liihafiil al-'lhhafut, trans S. van den Hcrgh (London, 1954), pp. 89. 96. 



4 



Introduction 



existence of anything having the first of the three specifications — the attribute of 
being the cause of the universe — which the proponents of proofs from eternity 
themselves deem requisite for the deity. 10 

Besides proofs for the existence of God from one or the other premise, from 
the premise of the eternity of the world or the premise of creation, proofs were 
also advanced with neither eternity nor creation as a premise. In some instances, 
creation is established in the course of proving the existence of God; a single 
train of reasoning arrives at both the existence of God and the creation of the 
world." In other instances, the issue of eternity and creation is nowise touched 
on. 12 Proofs of the latter sort avoid taking a position on the presence or absence 
of volition in the deity. Nevertheless, they cannot avoid taking a position on the 
specifications for the deity. By professing to be genuine proofs of the existence 
of God without demonstrating that God possesses will, they tacitly affirm that 
will is not a requisite specification. 

Still another procedure for proving the existence of God was in evidence. Ibn 
Tufayl and Maimonides, as well as the Scholastic philosopher Thomas Aquinas, 
do not themselves subscribe to the eternity of the world, yet they put forward 
proofs of the existence of God resting on the premise of eternity. They demon- 
strate the existence of God on two parallel, alternative tracks, on both the hypoth- 
esis of eternity and the hypothesis of creation; and proofs resting on the premise 
of eternity serve them hypothetically, as a means for establishing the existence 
of God on one of the two conceivable alternatives. The world, so their reasoning 
goes, either is eternal or had a beginning. Should the world be eternal, the 
Aristotelian proof from motion and other proofs from the premise of eternity 
establish a first incorporeal cause, who is the deity; whereas if the world should 
not be eternal, its having come into existence permits the immediate inference 
of a creator. In either event, the existence of God is established. Ibn Tufayl chose 
the procedure outlined in order to refrain from any stand whatsoever on the issue 
of eternity and creation. He explains that the issue is unrcsolvablc and that the 
existence of God will have been demonstrated only if shown to follow from both 
the hypothesis of the eternity of the world and the hypothesis of creation. 13 The 
rationale of Maimonides and Aquinas was different. They do take a stand on the 
issue of eternity and creation, and advocate the latter. They nonetheless wish to 



10 Ghazali, Tahafut al-Falasifa, III, §§17, 28; X, §1; English translation, pp. 96, 102, 250. The 
contention that eternal causation is nonsensical is found elsewhere; cf. below, pp. 190, 193, 210. 
"Below, pp. 149-150, 151, 188, 190, 387. 

l2 The notable example is Avicenna's proof, below, Chapter IX. Kalam writers subsequent to 
Avicenna commonly advance versions of Avicenna's proof side by side with the traditional Kalam 
arguments. 

"Hayy ben Yaqdhan, ed, and trans. L. Gauthicr (Beirut, 1936), Arabic text, pp. 81-86; French 
translation, pp. 62-65; English translation with pagination of Ihc Arabic indicated: Hayylbn Yaqzan, 
trans. L. Goodman (New York, 1972). 



Introduction 



5 



avoid linking the more fundamental and less problematic doctrine of the existence 
of God to the less fundamental and more problematic doctrine of creation. They 
therefore leave the question of creation open while demonstrating the existence 
of God and, like Ibn Tufayl, prove the existence of God on parallel tracks, on 
both the hypothesis of eternity and the hypothesis of creation. 14 Once having 
shown that God exists whether or not the world is eternal, they return, however — 
unlike Ibn Tufayl— to the issue of creation and eternity; and Maimonides, for his 
part, offers "arguments," acknowledged by him to be less than apodictic dem- 
onstrations, for creation. 15 In espousing creation, Maimonides and Aquinas clearly 
espouse, as well, the presence of volition in God. But by postponing the subject 
of creation until after proving the existence of God, they tacitly affirm that volition 
is not integral to the concept of the deity, that a proof of the existence of a single 
incorporeal cause, albeit a necessary cause, would constitute a genuine proof of 
the existence of God. 

Various procedures for proving the existence of God are, in sum, discernible 
in medieval Islamic and Jewish philosophy, sonic of which do, whereas others 
do not, require that the issue of eternity and creation be settled before a proof of 
the existence of God can be accomplished. The existence of God might be proved 
through the premise of the world's eternity. In proofs of the sort, the deity is 
explicitly or implicitly defined as an uncaused cause, incorporeal, and one. The 
existence of God might be proved through the contrary premise, creation, the 
attendant conclusion being a deity possessed of will. Ghazali, who expatiates 
upon the import of proofs based on the premise of creation, stresses that volition 
is integral to an adequate concept of the deity, that unless a proof establishes a 
first cause possessed of will, it is not a genuine proof of the existence of God. 
The existence of God, creation, and hence the attribute of will, might all be 
established by means of a single train of reasoning. Here, it is not clear whether 
volition is viewed as integral to the concept of God. The existence of God might 
also be proved utilizing neither the premise of eternity nor the premise of creation. 
The implication now would plainly be that whether or not God docs possess will, 
volition is not part of the irreducible concept 'of the deity. Finally, a proof from 
the premise of eternity might be employed by philosophers who do not themselves 
subscribe to the premise. The proof from eternity would serve a hypothetical 
function and would be supplemented through a parallel proof from creation. 
Again, the implication would be that the irreducible concept or the deity docs 
not contain volition. 



"Maimonides, Guide, I, 71; II, 2; Aquinas, Summa contra Gentiles, I, 13. Also cf. Aquinas, 
Commentary on Physics, VIII, §970, and Commentary on Metaphysics, XII, lectio 5 (end). 

"Maimonides, Guide, II, 19; 22. Aquinas, Summa Theologiae, I. q. 46, arts. 1, 2; Summa 
contra Gentiles, II, chap. 38. Also cf. A. Maicr, "Problem des aktucll Unendlichen," Ausgehendes 
Miltelalter, Vol. I (Rome, 1964), p. 48. 



6 



Introduction 



In a later chapter it will be seen that medieval Islamic and Jewish arguments 
for the existence of God are, in the main, cosmological; tclcological arguments 
are also found; and no argument is ontological. 16 

2. The present book 

The chapters to follow examine medieval Islamic and Jewish arguments for 
eternity; responses to the arguments for eternity on the part of the proponents of 
creation; medieval Islamic and Jewish arguments for creation; responses thereto 
by the proponents of eternity; proofs for the existence of God resting on the 
premise of creation, the premise of eternity, or neither premise; and refutations 
of proofs for the existence of God. 

Medieval Islamic and Jewish philosophy will be treated here as a single philo- 
sophic tradition. Treating the several branches of medieval philosophy — Islamic, 
Jewish, and Christian — in conjunction with one another is surely justified con- 
sidering the extent to which they draw sustenance from the same, or similar, 
sources and are animated by the same spirit. 17 From the standpoint of medieval 
Jewish philosophy, treating Islamic and Jewish philosophy conjointly has an 
additional justification, inasmuch as Jewish philosophy is rooted in Islamic phi- 
losophy and cannot be properly understood in isolation from it. The reverse is 
of course not true: Medieval Islamic philosophers barely knew of the existence 
of the Jewish philosophers. Yet when viewed historically, Jewish philosophy docs 
shed light on medieval Islamic philosophy in a number of areas, notably in two. 
For the early period, Jewish thinkers who read and wrote Arabic, especially 
Saadia, complement the available Islamic material and help delineate the begin- 
nings of Arabic philosophy; and for a later period, Hebrew sources are an inval- 
uable aid in the study of Averroes. Many of Averroes' writings are preserved 
exclusively in Hebrew, and Hebrew commentaries furnish the best, and often the 
sole, access to the meaning of even those preserved in Arabic. 

I have not tried to take account of each and every argument for eternity, 
creation, and the existence of God, which might be unearthed in medieval Arabic 
and Hebrew literature, but I have tried to take account of every argument for 
those doctrines which is of a philosophic character. Kalam reasoning conse- 
quently falls within the scope of the present work when it appears to mc to be 
philosophic, but has been disregarded when it docs not. The difficulty in drawing 
up a satisfactory definition of philosophy always lies in the demarcation of the 
precise boundary between philosophy and other domains. Still, any thoughtful 
acceptation of the term will distinguish philosophy from pure theology, from 
speculation whose underlying premises are founded entirely on religious faith, 
and will also exclude from the domain of philosophy enunciations that fall below 
some minimal level of plausibility. The Kalam arguments that I have disregarded 

"■Below, pp. 214-216. 

"C(. H. Wolfson. Philo (Cambridge, Mass., 1948), chap. 14. 



Introduction 



7 



arc of a purely theological character or so unsubstantial that they cannot, even 
with magnanimity, be taken seriously. 18 

My discussion purports to be exhaustive as regards Arabic and Hebrew argu- 
ments; that is, I have undertaken to examine every medieval Arabic and Hebrew 
philosophic argument for eternity, creation, and the existence of God. Arguments 
in the two literatures which are not discussed were either judged to be nonphilo- 
sophic, were subsumed under other arguments that are discussed elsewhere in 
the book, 19 or were simply overlooked. In a number of instances 1 have pursued 
the penetration of Islamic and Jewish arguments into medieval Christian philos- 
ophy, and in a few instances into modern European philosophy. There, though, 
1 make no pretense at exhaustiveness, and the citations are of a kind that are ready 
at hand in obvious primary and secondary works. 

The discipline to which the present study belongs is what I would venture to 
call the history of philosophic ideas. I am attempting to trace the history of 
philosophic constructions rather than reproduce the complete system of any indi- 
vidual philosopher or philosophic school. My concern is chiefly with the history 
of proofs — of eternity, creation, and the existence of God; but also worthy of 
attention is the history of the components from which the proofs are fashioned. 
The starting point both for the history of the proofs and the history of their 
components is, with rare exceptions, Aristotle. That is hardly surprising; the 
importance of Aristotle for medieval philosophy is common knowledge. The 
direction in which the Aristotelian conceptions developed in the Middle Ages 
was, however, often determined by the late Greek philosophers, and their impor- 
tance for the Middle Ages is far from common knowledge. In much of the material 
to be examined, Proclus (5th century) and, in greater measure, John Philoponus 
(6th century) are responsible for the direction in which Aristotelian conceptions 
developed. The significance for medieval Islamic and Jewish thought of Proclus, 
one of the last Greek Neoplatonists, and of Philoponus, a Christian and sometime 
commentator on Aristotle, is a subordinate theme of the present book. 

The proofs for eternity, creation, and the existence of God are admirably suited 
for an organized presentation. They arrange themselves spontaneously and grace- 
fully into chapters and chapter sections. Unfortunately, the arrangement into 
chapters and sections according to the proofs tends to obscure the history of the 
components from which the proofs are fashioned. The components wend their 
way through one proof after another. A principle or conception may originate in 
Aristotle; undergo development at the hands of Proclus or Philoponus, or both; 

l8 I have, for example, passed over the argument for crealion offered by Ash ' ari, K. al-Luma ' , 
§3 (in The Theology of al-Ash' art. ed. and trans. R. McCarthy (Beirut, 1953)), and also recorded 
by Shahrastani. K. Nihava al-Iqdam, ed. A. Guillaumc (Oxford and London, 1934), p. 12. H. 
Wolfson, Philosophy of the Kalam (Cambridge, Mass., 1976). pp. 382-383, terms it the "argument 
from the analogy of things in the world." It is simply an expansion of Quran 23:12-14. 

"For example, some of the Kalam arguments for creation listed by Wolfson, Philosophy of the 
Kalam, p. 374, will be treated here. Chapters IV, V, and VI, as variations of other arguments. 



8 



Introduction 



and repeatedly cross linguistic, confessional, and school boundaries — as well as 
the chapter headings of the present book — while traveling down fifteen centuries 
or more. In crossing boundaries, the principle may remain imbedded within the 
context with which it was originally associated, but may also be transformed and 
reembodied in a new context. And a principle or conception originally formulated 
by a given party may be converted by the opposing party into ammunition for 
attacking the basic tenets of the former. 20 Particularly far-reaching examples will 
be offered in Appendix A. It traces the peregrinations of two Aristotelian prin- 
ciples — the principle that an infinite number is impossible, and the principle that 
a finite body can contain only finite power — which make their appearance in 
several chapters but do not lend themselves to a complete treatment at any single 
place in the body of the book. 21 

Medieval Islamic and Jewish arguments for eternity and responses made to 
them by the adherents of creation are discussed in Chapters II and III. Chapters 
IV and V treat a group of arguments for creation which can be linked with John 
Philoponus, together with refutations of the arguments. Chapters VI and VII 
explore two styles of reasoning, argumentation from the concept of particulari- 
zation and teleological argumentation, which were used in the formulating of 
proofs for both creation and the existence of God. Chapter VIII analyzes the 
Aristotelian proof of the existence of God as a prime mover, together with Hasdai 
Crescas' refutation. Chapter IX examines Avicenna's proof of the existence of 
God as a being necessarily existent by virtue of itself, and Chapter X analyzes 
Averroes' refutation of Avicenna's proof. Chapter XI discusses a family of proofs 
for the existence of God, all resting on the principle that an infinite regress of 
causes is impossible; refutations of the proofs are treated in the same chapter. 
Chapter XII deals with the history, subsequent to Avicenna, of arguments for the 
existence of God as a necessary being. 

All the translations are mine. 



This occurs in connection with both principles discussed in Appendix A. 
2 'Other principles appearing repeatedly — and employed by different parties, each for its own 
purpose — are the principle that every possibility must eventually be realized; and the principle that 
whatever is generated is destructible, and vice versa. 



II 



Proofs of Eternity from the 
Nature of the World 



1. Proofs of eternity 

The writings of medieval Islamic and Jewish philosophers reveal stock proofs 
of the eternity of the world, proofs that recur repeatedly through the centuries 
with little change. Proofs of the eternity of the world are, of course, advanced 
by philosophers espousing eternity. But they are also recorded by philosophers 
who, on the contrary, espouse creation; and since the advocates of creation were 
more numerous than the advocates of eternity, more citations of the standard 
proofs for eternity can be gleaned from the writings of the former than from the 
writings of the latter. When the advocates of creation record proofs of eternity, 
their purpose, naturally enough, is to prepare the way for a refutation. In that, 
John Philoponus may well have served as the model; Philoponus had painstakingly 
refuted all the arguments for eternity which he had discovered in Aristotle and 
Proclus, and his refutations of Aristotle and Proclus together with his own argu- 
ments for creation were known to, and used by, the Islamic and Jewish philos- 
ophers. 1 Not only were the proofs for eternity standardized; the refutations were, 
as well. A repertoire of stock refutations was thus arrayed against a repertoire of 
stock arguments. And the refutations, in their turn, often elicited surrejoinders 
from the proponents of eternity. This chapter and the next examine the philosophic 
proofs for eternity appearing in the writings of medieval Islamic and Jewish 
advocates of eternity or creation; rebuttals of the proofs for eternity; and surre- 
joinders to the rebuttals. 

Ancient and medieval proofs of eternity do not all have the same aim. The 
more comprehensive seek to establish that the world has existed from eternity in 
the form in which it exists today. But less comprehensive proofs, too, arc in 
evidence. They seek to establish the eternity of the matter of the world, leaving 
open the possibility that the form of the world is not eternal, that the world was, 

'Examples are offered in this and the following three chapters. 



9 



10 



Proofs of Eternity from Nature of World 



in other words, created out of a prccxistcnt matter. Unfortunately, both adherents 
of eternity and adherents of creation often fail to state explicitly what a given 
proof is intended to accomplish; and it sometimes even happens that proofs 
ostensibly of the eternity of the world as a whole argue, in fact, for nothing more 
than the eternity of matter. 

As will be seen in later chapters, the doctrine of creation is likewise supported 
by more comprehensive and less comprehensive proofs. A more comprehensive 
proof of creation seeks to establish the creation of the world ex nihilo, whereas 
a less comprehensive proof seeks only to establish the creation of the form of 
the world, leaving open the possibility that the matter of the world is eternal. A 
less comprehensive proof of eternity, that is, a proof of the eternity of matter, is 
plainly compatible with a less comprehensive proof of creation, a proof solely 
of the creation of the world in its present form. Accordingly, a proof of the 
eternity of matter is compatible as well with volition in the deity; for creation in 
any mode, even creation from a preexistent matter, would constitute a voluntary 
act on the part of the creator. 

Maimonides drew a dichotomy that is highly helpful, if not indispensable, for 
any analysis of proofs for eternity. The proofs of eternity deserving consideration, 
he explained, fall into two categories. There are, on the one hand, arguments 
formulated by Aristotle which take their departure "from the world." On the other 
hand, there are arguments "extracted" by subsequent philosophers "from Aris- 
totle's philosophy," these, in contrast to the previous category, taking their depar- 
ture "from God." 2 The distinction between proofs of eternity from the nature of 
the world and proofs from the nature of God does not seem to have been artic- 
ulated before Maimonides, but almost every medieval Islamic and Jewish phi- 
losopher who treats the issue of eternity does adduce proofs belonging to both 
categories. If Maimonides' dichotomy were to be collated with the distinction 
between proofs of the eternity of the world in its form as well as its matter and 
proofs of the eternity of matter alone, the result would be as follows: Some proofs 
from the nature of the world, it turns out, seek to demonstrate the eternity of the 
universe — or, to be more specific, the eternity of the physical universe 3 — in its 
entirety, whereas others merely have the aim of demonstrating the eternity of 
matter; 4 proofs of eternity from the nature of God, however, are invariably of 
the more comprehensive kind and aim at demonstrating the eternity of the uni- 
verse in its entirety. 

2 Guide to the Perplexed, II, 14. The dichotomy reappears in Albertus Magnus, Bonavcnture, 
Aaron ben Elijah, and Crcscas. Aquinas in (he Summa contra Gentiles, II, 32-34, subdivides proofs 
from the nature of the world into two categories and thereby reaches a trichotomy. Isaac Abravanel, 
who knew Aquinas' works, has a similar trichotomy in Mif'alot (Venice, 1592), IV- VI. 

'Proofs from the nature of the physical world can lead, naturally enough, only to the eternity 
of the physical world or to the eternity of the underlying matter of the world. Proofs from the nature 
of the deity would lead to the eternity of the entire universe, nonphysical as well as physical. 

"See below, pp. 29-30. 



Proofs of Eternity from Nature of World 



11 



Maimonides' identification of Aristotle as the author of the proofs for eternity 
from the nature of the world is not inaccurate. The medieval Islamic and Jewish 
proofs in that category either are borrowed directly from Aristotle or are adap- 
tations of arguments he put forward. Maimonides' description of proofs from the 
nature of God as having been "extracted . . . from Aristotle's philosophy," in 
other words, as having been fashioned out of Aristotelian components, is less 
exact; for one of the three proofs in this category is animated by a Neoplatonic 
theme that Aristotle would have found uncongenial. The main source or channel 
through which proofs for eternity from the nature of God reached the Middle 
Ages was apparently Proclus. 5 

Heeding Maimonides' dichotomy, I deal in the remainder of the present chapter 
with proofs of eternity from the nature of the world, and in the next chapter with 
proofs of eternity from the nature of God. Besides those discussed, arguments 
for eternity are occasionally referred to in the writings of the proponents of 
creation which are so flimsy that they can hardly be regarded as more than straw 
men; they cannot conceivably have been advanced by intelligent, self-respecting 
proponents of the eternity of the world. 6 Arguments of the sort have been dis- 
regarded here except insofar as they can be related to serious philosophic proofs. 

In classifying proofs of eternity, it is sometimes difficult to decide whether a 
proof should be assigned to one general category or the other, to the category 
that reasons from the nature of the world or to the category that reasons from the 
nature of God. 7 And when enumerating separate and distinct proofs within each 
category, arbitrariness is unavoidable. Through the centuries the arguments sub- 
divided and underwent variations; and should all subdivisions and variations be 
listed as separate proofs, the total number swells. For example, arguments for 
eternity which Maimonides listed as single proofs 8 were later subdivided by 
Crescas into two or three proofs, with the result that Crescas' list is several times 
as long as Maimonides'. 9 Aquinas, by enumerating subdivisions and variations, 
was able in one work to draw up a list of over twenty-five philosophic proofs for 
eternity. 10 Whereas some philosophers thus subdivide the arguments, others com- 
bine them. Proofs that can stand independently may coalesce to form a single 
complex chain of reasoning. Or else, what was originally an independent proof 
for eternity may be advanced by a philosopher not as such, but as a surrejoinder 

'See below, p. 51. 

"Sec, for instance, Ibn Hazm, K. al-fasl fi al-Milat (Cairo, 1964), I, p. 10(5). Spanish trans- 
lation: Abrnhdzam de Cordoba y su Historia Crilica de las Ideas Religiosas. trans. M. Asm Palacios, 
Vol. II (Madrid, 1928), p. 99. 

7 For example, the argument from motion, which is discussed in this chapter, blends into the 
argument from the unchangcability of the cause of the universe, which is discussed in the next chapter. 

'Guide, II, 14. 

"Or ha-Shem, III, i, I . Similarly, Abravanel, Mifalot, IV, 3; V, 1 ; VI, I . 
w De Potcntia, q.3, art. 17. Aquinas lists, in fact, thirty arguments for eternity, but some are 
scriptural and not philosophic, and some arc duplicates. 



12 



Proofs of Eternity from Nature of World 



to the refutation of another proof of eternity." 1 found that the requirements of 
exposition would be best served by distinguishing six separate proofs of eternity 
from the nature of the world and three separate proofs from the nature of God. 
Alternative enumerations would undoubtedly be equally justified. 

The medieval advocates of eternity who will be cited here arc: the authors of 
a corpus of Arabic writings (precise dates unknown) attributed to Alexander of 
Aphrodisias, Avicenna (980-1037), Abu al-Barakat (d. ca. 1160), Averroes (1 126- 
1198), and Moses Narboni (d. 1362). In addition, two philosophers argue for 
the eternity of matter, although they espouse the creation of the world in its 
present form; they are Abu Bakr b. Zakarlya RazI (ca. 864-925) and Gersonides 
(1288-1344). The medieval advocates of creation ex nihilo who will be cited 
are: Saadia (892-942), Baqillam(d. 1013), 'Abd al-Jabbar(d. 1025), IbnHazm 
(994-1064), the authors of the Jabir corpus (10th century), Juwaynl (1028- 
1085), Bazdawl (Pazdawl) (d. 1099), Ghazali (1058-1 111), ShahrastanT (1086- 
1153), Maimonides (1135-1204), Fakhr al-Din al-RazI (1 149-1209/10), Amid! 
(1156-1233), TusI (d. 1274), Albertus Magnus (1206-1280), Bonaventurc (1221- 
1274), Thomas Aquinas (1225-1274), IjT (d. 1355), 12 Aaron ben Elijah (d. 1369), 
Hasdai Crescas (1340-1410), Joseph Albo (1380-1444), and Isaac Abravanel 
(1437-1508). A final writer to be cited, Leone Ebreo (Judah Abravanel) (d. 1535), 
does not state clearly whether he endorses creation from a preexistcnt matter or 
creation ex nihilo. 

2. Proofs of eternity from the nature of the physical world 

Six proofs for eternity from the nature of the world can, as was mentioned, 
be distinguished; and some of the six undergo variations. In every instance Aris- 
totle is either the immediate or ultimate source. 

The first four of the six proofs disclose an identical structure. To begin, they 
all proceed indirectly, establishing their own thesis by focusing on the opponent's 
position and showing it to be untenable: They argue that the world must be eternal 
since creation is impossible. Indirect reasoning is far from unique to them, how- 
ever. The fifth proof from the nature of the world also reasons indirectly. Much 
of the argumentation in the second category, where eternity is proved from the 
nature of God, proceeds in the same fashion. And the advocates of creation too 
had a predilection for indirect reasoning; proofs for creation, as will appear in 
later chapters, typically focus on the thesis of eternity and argue that it is untenable. 

What is unique in the first four proofs for eternity is the manner in which they 
establish the untenability of creation and thereby indirectly establish the eternity 
of the world. They all contend that the laws of nature — or, to be more accurate, 

"For example, below, pp. 75-76. 

,2 I have not distinguished between the text of Iji and the commentary of JurjanI, which is 
interwoven with it. 



Proofs of Eternity from Nature of World 



13 



the laws of Aristotelian physics — are such that for creation to have taken place, 
something would already have had to exist similar to what was supposedly coming 
into existence for the first time. In order that matter should have come into 
existence, matter would already have had to exist; and in order that a world 
should come into existence, a world would already have had to exist. Hence the 
assumption of an absolute beginning of matter or an absolute beginning of the 
world is self-contradictory, and the world, or matter, must exist from eternity. 
I turn now to the individual proofs. 

(a) The argument from the nature of matter 

In the course of an exhaustive analysis of the nature of matter, Aristotle argued 
that the underlying matter of the universe must be eternal, and his procedure 
reveals the pattern just outlined. He established the eternity of matter by showing 
the creation of matter to be untenable; and his grounds for the untenability of the 
creation of matter arc that matter could only have come into existence from an 
already existent matter. Aristotle reasoned: Everything that comes into existence 
does so from a substratum. 13 If the underlying matter of the universe came into 
existence, it also would come into existence from a substratum. But the nature 
of matter is precisely to be the substratum from which other things arise. Con- 
sequently, the underlying matter of the universe could have come into existence 
only from an already existing matter exactly like itself; and to assume that the 
underlying matter of the universe came into existence would require assuming 
that an underlying matter already existed. The assumption is thus self-contradic- 
tory, and matter must be eternal. 14 The argument that the underlying matter of 
the universe must be eternal since matter could only come into existence from 
an already existing, prior matter is adduced by the adherents of eternity in defense 
of their position, and it is recorded by adherents of creation who have in view a 
subsequent refutation. The adherents of the eternity of matter adducing the argu- 
ment arc Avicenna," Averroes, 16 and Gersonides; 17 the adherents of creation 
recording it arc Maimonides, 18 Albertus Magnus, 1 '' Aquinas, 20 IjT, 2 1 Aaron ben 

"Physics I, 7. 

"Physics I, 9, 192a, 29-31. 

[i Shifa': Ilahiyat, cd. G. Anawati and S. Zayed (Cairo, 1960), p. 376. 

i( " Epitome of Physics, in Rasa'il Ibn Rushd (Hyderabad, 1947), p. 10; Middle Commentary on 
Physics (Oxford, Bodleian Library, Hebrew MS. Ncubaucr 1380 = Hunt. 79), I, iii, 3 and 5; Long 
Commentary on Physics, in Aristotelis Opera cum Averrois Commentariis, Vol. IV (Venice, 1562), 
I, comm. 82; Long Commentary on Ue Caelo, in ibid., Vol. V, 1, comm. 22. 

"Milhamol ha-Shem (Leipzig, 1866). VI, i, 3, pp. 300, 302. In VI, i, 17, p. 364, Gersonides 
in effect accepts the argument. 

"Guide, 11, 14(2). 

'"Physics, in Opera Omnia, cd. A. Borgnet. Vol. Ill (Paris, 1890), VIII, tr. 1. chap. 11. 
2n Compendium of Theology, chap. 99. Cf. Summa Thcologiae, I, 46, I, obj. 3. 
2, Mawaqif (Cairo, 1907), VII, p. 228. 



( 

i 

\ 

, 14 Proofs of Eternity from Nature of World 

( Elijah, 22 Crescas, 23 Isaac Abravancl, 24 and Leone Bbrco. 25 In most instances, 

(he force of the argument is sharpened. Matter, in the Aristotelian physical scheme, 

never exists in actuality devoid of all form, and the generation of any object 
/ therefore starts not merely with a previously existing matter but with a previously 

existing compound of matter and form. 26 Generation is the process wherein the 
'• compound of matter and form receives an additional form or exchanges its own 

/ form for another. The sharpened medieval version of the argument for the eternity 

of matter accordingly runs: The underlying matter of the world could only have 

come into existence from an already existing matter. Matter, however, never exists 
( devoid of form, but solely in a compound, together with form. To assume an 

absolute beginning of the underlying matter of the universe would, then, not 
( merely require assuming an already existent matter before matter existed. It would 

( require assuming that before matter existed, there already existed a full-blown 

compound of matter and form. 27 
( The heart of the foregoing argument or arguments clearly is the principle that 

( whatever comes into existence does so from a substratum. Once the principle is 

accepted, discussion is virtually precluded; for if whatever comes into existence 
( does so from an already existing substratum or matter, matter obviously is eternal. 

( As for the critical principle, Aristotle in one passage supported it inductively. 

We can, he wrote, "always" observe "something underlying, from which the 
( generated object comes, plants and animals, for example, [coming] from seed." 28 

( Elsewhere, though, Aristotle treated the impossibility "that generation should 

take place from nothing" as self-evident. 29 
'■. In the Middle Ages, the presupposition upon which the argument for the eterni- 

/ ty of matter rests received frequent attention. Generally it was adherents of the 

opposing position whose attention was roused, their object being to expose the 



( argument's feeble foundation; but an exception to the generalization is Zakariya 

RazT. RazT, an advocate of the eternity of matter although not of the eternity of 
the world, is reported to have maintained the eternity of matter on the grounds 
( that "intellect" rejects the coming into existence of something from nothing. 30 



^ In effect, that is to say, RazI regarded the presupposition of the argument for the 

( 

22 '£s Hayyim, ed. F. Delitzsch (Leipzig, 1841), chap. 6. 
23 Or ha-Shem, III, i, 1. 
24 Mifalot. V, 1(1). 

2s Dialoghi cl'Amore (Ban, 1929), p. 237; Hebrew translation: Wikkuah 'al ha-Ahaba (Lyck, 
1871), p. 55b; English translation: The Philosophy of Love, trans. F. Friedcberg-Secley and Jean H. 
Barnes (London, 1937), pp. 278-279. 
| 26 De Generatione II, I, 329a, 24-26; Physics IV, 7, 214a. 15. 

27 Such is the implication in Averroes' version. The argumentation is explicit in Maimonides, 
; Alberlus, Iji, Aaron ben Elijah, Crescas, and Abravancl. 

2> Physics I, 7, 190b, 3-5. 
; ( 29 'Metaphysics III, 4, 999b, 8; see Ross's note, ad locum. 
: 30 RazI, Opera Philosphica, ed. P. Kraus (Cairo, 1939), I, p. 221. 

( 
I 



Proofs of Eternity from Nature of World 



15 



eternity of matter as self-evident. Aquinas, Aaron ben Elijah, and Crescas, ail 
advocates of creation ex niliilo, similarly record arguments for the eternity of 
matter which rest on a supposedly self-evident proposition. Aquinas reports an 
argument for "eternity" — which is in fact an argument for the eternity of mat- 
ter — resting on the "universal opinion of philosophers that nothing is made from 
nothing." 31 Aaron and Crescas likewise refer to an argument for the eternity of 
matter resting on the premise that — in Aaron's words — "something cannot be 
generated from nothing," or — in Crescas' words — "the generation of a body out 
of nothing is absurd." 32 

Among Kalam writers and writers affiliated with the Kalam, the belief not 
merely in the eternity of matter but also in the eternity of the world is commonly 
traced to an induction or analogy, one that the Kalam, of course, deemed improp- 
erly drawn. 33 Saadia, for example, avers that the "strongest argument" of the 
adherents of the eternity of the world in its entirety (al-dahr) is the circumstance 
that an absolute beginning of the processes of nature has never been witnessed. 34 
The Jabir corpus wrestles with, and endeavors to refute, two parallel contentions, 
one to the effect that the world must be eternal since nobody ever observed the 
world's coming into existence, and the other to the effect that every man must 
have been born of woman since man was never observed not to be born of 
woman. 35 Ibn Hazm represents the adherents of eternity (dahriya) as reasoning: 
"We do not sec anything coming into existence except from something or in 
something, and whoever maintains the contrary is maintaining what neither is 
now observed nor has ever been observed." 36 'Abd al-Jabbar records a proof 
possibly containing an echo of the passage in which Aristotle cited the fact that 
"plants and animals [come] from seed." 37 According to 'Abd al-Iabbar, the advo- 
cates of eternity reasoned: "We do not find a hen except from an egg or an egg 
except from a hen, and things must therefore always have been so. That implies 
the eternity of the world." 38 Juwaynl's account is similar to 'Abd al-Jabbar's. He 
describes the advocates of eternity (dahriya) as arguing: "Since we never observe 
a hen except from an egg or a man except from the sperm of male and female, 
the judgment that such takes place must be extended to what is hidden from us 
[in past time]." 39 Bazdawi reports an argument for the eternity of matter resting 



3, Sttmtna contra Gentiles, II, 34(1). 

}1 'Es Hayyim, chap. 6; Or ha-Shem, III, i, 1 . 

"See below, pp. 30. 34. 

34 K. al-Amanat wa-II' liqadat, ed. S. Landaucr (Leiden, 1880), I, 3(10). p. 63. English trans- 
lation, with pagination of the Arabic indicated: Book of Beliefs and Opinions, trans. S. Rosenblatt 
(New Haven, 1948). 

35 Jabir ibn Hayyan, Tcxtes Choisis, ed. P. Kraus (Cairo, 1935). p. 422. 

36 AT. al-Fasifi al-Milal, I, p. 9; Spanish translation, II, p. 94. 

"Above, n. 28. 

'"Sharh al-Usul (Cairo, 1965), p. 1 17. See below. Chapter V, n. 162. 
39 A\ al-Shamil (Alexandria, 1969), p. 224. 



16 



Proofs of Eternity from Nature of World 



on the circumstance that "we never have seen' anything created from nothing," 
and concluding "that the world was not created from nothing." 40 Aaron ben 
Elijah— who, as already seen, refers to an argument for the eternity of matter 
based on the supposedly self-evident "premise . . . that something cannot be 
generated from nothing" 41 — also gives what seems to be an argument by induc- 
tion: Since what is generated is generated "only from something else ... and 
only from something specifically [adapted to the resulting product] . . .it follows 
a fortiori that nothing can be generated out of nothing." 42 Finally, Isaac Abravanel 
records two closely related arguments for the eternity of matter, one in almost 
the same language as Aaron, and the other explicitly labeled as "inductive." The 
latter begins with an examination of "particular things, whether substances or 
accidents, which come into existence"; it discovers that whatever comes into 
existence does so from "from something, not from nothing"; and it arrives at the 
"judgment that everything comes into existence from something, and nothing 
can come into existence from nothing." 43 

(b) The argument from the concept of possibility 

In the Aristotelian physical system, possibility and matter are closely related 
concepts, matter being the locus of potentiality, or possibility. 44 The relationship 
between the concepts led to the development of a proof for the eternity of matter 
from the concept of possibility, alongside the proof from the nature of matter, 
examined in the preceding paragraphs. The germ of the proof from the concept 
of possibility can be discovered easily enough in Aristotle. In one passage, for 
example, Aristotle established that the process of coming into existence "nec- 
essarily" requires "the prior presence of something existent potentially [or: pos- 
sibly], but not existent in actuality" ; 45 it follows that nothing whatsoever, includ- 
ing matter, can come into existence from absolute nothingness. In the Middle 
Ages, the proof takes a distinctive cast, however, and the earliest philosopher to 
whom I could trace the new and distinctive formulation is Avicenna. Considering 
that the concept of possibility was a central concern of Avicenna 's, 46 he very 
likely is the author of the argument as it is found in the Middle Ages. 

Avicenna lays down the proposition that prior to a thing's coming into actual 
existence, its existence must have been possible; were its existence necessary, he 
explains, the thing would already have existed, and were its existence impossible, 
the thing would never exist. 47 The "possibility (imkdn) of the existence" of a 



40 Bazdawi (Pazdawi), K. Usui al-Din, ed. H. Linss (Cairo, 1963) p. 16. 
"'Above, n. 32. 
42 'Es Hayyim, chap. 6. 
^Mifalot, IV, 3(1,2). 

**Cf. Metaphysics XIV, 1, 1088b, 1. Potentiality and possibility both translate the Greek 5wa|xi.s. 
45 De Generation I, 3, 317b, 16-17. 
46 See below, Chapter IX. 

47 Cf. Aristotle, De Generatione II, 9, 335a, 32 ff. 



Proofs of Eternity from Nature of World 



17 



thing must, moreover, in some sense have its own existence. For if the possibility 
were completely "nonexistent," it could not legitimately be spoken of as "being 
prior" to the thing's actual existence, and everyone will surely acknowledge that 
the possibility is prior. Mere possibility of existence is plainly not a substance. 
It can only belong to the class of entities that are "present in a subject." Thus 
whenever anything comes into existence, the possibility of its existence must 
previously have subsisted in a subject. The "possibility (imkiin) of existence" 
may also be termed the "potentiality (t/uwa) of existence"; 48 and the subject in 
which possibility or potentiality is found is called "matter" (madda, hayula). 
Whenever anything comes into existence, the possibility or potentiality of its 
existence must, then, previously have subsisted in an already existent subject, or 
matter. But, Avicenna concludes, if an already existent matter must precede 
everything coming into existence, clearly nothing, including matter, can come 
into existence ex nihilo, that is, from absolute nothingness. An absolute beginning 
of the existence of matter is impossible. 49 

The argument for the eternity of matter from the concept of possibility is cited 
in Ghazali's critique of Avicenna's philosophic system, 50 in ShahrastanT's account 
of Avicenna's philosophy, 51 in Fakhr al-DIn al-RazI, 52 Averroes, 53 Maimon- 
ides, 54 Amidi, 55 Albertus Magnus, 56 Aquinas, 57 Gersonides, 58 Ijl, 59 Aaron ben 
Elijah, 60 and Isaac Abravanel. 61 

(c) The argument from the nature of motion 

In Physics VIII, 1 , as a preliminary to his demonstration of the existence of a 
first mover of the universe, Aristotle undertook to prove the eternity of motion. 
His proof contains several strands, and they were put to various uses in the Middle 
Ages. 



4s 8ui>ap.is underlies both imkan and quwa. 

A9 Najat (Cairo. 1938), pp. 219-220; cf. Shifa': llahiyat, pp. 177-178. 

'"Tahafut al-l-alasifa, ed. M. Bouygcs (Beirut. 1927), I. §§81-82; English translation in Aver- 
roes' Talutfut al-Tahafut, trans. S. van den Bcrgh (London. 1954). p. 57. 

5 'K. al-Milal wa-l-Nihal, ed. W. Cureton (London, 1842-1846), p. 371. 

52 Muhassal (Cairo, 1905). p. 9\;K. al-Arha'in (Hyderabad, 1934), p. 49. 

"Tahaftit al-Tahafut, ed. M. Bouygcs (Beirut, 1930), 1, pp. 69, 74, 100; English translation, 
with pagination of the Arabic indicated; Averroes' Tahafut al-Tahafut, trans. S. van den Bergh 
(London, 1954). 

"Guide, II, 14(4). 

"Ghiiya al-Maram (Cairo, 1971), p. 267. 
^Physics. VIII, tr. 1, chap. 11. 

"Summa Theologiae, I. 46. 1 , obj. 1 ; Summa contra Gentiles, II, 34(3); De Potentia, q. 3, art. 
17(10). 

^Milhamot ha-Shem, VI, i, 17, p. 365; he accepts the argument. 
i9 Mawaqif, VII, p. 230; the argument is not fully developed. 
m 'Es Hayyim. chap. 6. 
"Mif alot, IV, 3(3). 



18 



Proofs of Eternity from Nature of World 



Aristotle's proof ran: Motion— taking the term in the broad sense that includes 
change of all types 62 — obviously can occur only if an object undergoing motion 
is "present" (vTrdpxeiv). 63 If an absolute beginning of motion should be assumed, 
the object to undergo the first motion must either (i) have come into existence, 
or (ii) have been eternal. That is to say, an absolute beginning of motion can be 
construed in one of two ways: (i) The world came into existence and began to 
move; or (ii) the world existed in an eternal state of rest before beginning to 
move. The eternity of motion will be established when both alternatives are ruled 
out. 

Inasmuch as coming into existence is one type of motion, 64 alternative (i) 
asserts in effect that before the world performed its absolutely first motion, it 
had already performed another motion, namely the motion of coming into exis- 
tence. Alternative (i) asserts, therefore, that the absolutely first motion was not 
after all the first motion, which is a self-contradiction; and the proposition that 
the world came into existence and thereupon began to move is thus untenable. 65 
It is to be noted that Aristotle fails to explain why the absolutely first movement 
may not have been precisely the coming into existence of the physical universe. 
On such a theory, motion and the physical universe would have come into exis- 
tence together; the requirement that motion takes place only if an object undergo- 
ing motion is "present" would be met; and an absolute beginning of motion as 
well as of the world could be defended. The question is taken up by Thcmistius 66 
and Averroes. 67 

After Aristotle had, to his satisfaction, ruled out alternative (i)— an absolute 
beginning of motion in a world that likewise had a beginning— alternative (ii) 
remained. Alternative (ii) is the thesis that motion had an absolute beginning but 
the world is eternal, in other words, the thesis that the world existed in a state 
of rest for an eternity before starting to move. Aristotle found the thesis to be 
inadmissible for two reasons. 



62 Aristotle distinguished four sorts of motion or change, each of which takes place in a different 
category: change in the category of substance, that is to say, coming into existence or destruction; 
change in the category of quantity, that is to say. growth or diminution; change in the category of 
quality, that is to say, alteration; and change in the category of place. His statements are by no means 
consistent, however. See E. Zeller, Die Philosophic der Griechen, Vol. II, Part 2, (4th cd.; Leipzig, 
1921), pp. 389-390; W. D. Ross, Aristotle (London, 1953), pp. 82-83. 

"Aristotle bases the proposition both on common sense and on his definition of motion 

"See n. 62. 

"Physics VIII, 1, 251a, 8-20. 

^Themistius deals with the question indirectly. He has Aristotle contend not that the object to 
undergo motion must be "present" in order for motion to take place, but that it must be "previously 
present." Something would accordingly always have to exist prior to the occurrence of motion. Sec 
Thcmistius, Paraphrase of Physics, ed. H. Schenkel, Commentaria in Arislotelem Graeca, Vol. V/2 
(Berlin, 1900), p. 210. 

"See below, p. 21. 



Proofs of Eternity from Nature of World 19 

(a) Just as some factor must be posited as the cause of the motion of whatever 
undergoes motion, so too, Aristotle writes, must a factor be posited as the cause 
of the state of rest of whatever is at rest. If the world had been at rest before 
starting to move, something must have produced its state of rest. But producing 
the state of rest would itself be a motion. A motion would consequently have 
preceded the supposed absolutely first motion; and the assumption that the phys- 
ical universe existed in a state of rest from all eternity before undergoing an 
absolutely first motion embodies, like the previous alternative, a self-contradiction. 68 

Aristotle's thinking, again, has puzzling aspects. In the first place, once he 
laid down the rule that everything at rest must be set at rest he could have 
proceeded more simply. He could have argued that the thesis of a world in eternal 
rest is in itself self-contradictory because it implies a prior motion whereby the 
world was set at rest; and the supposedly eternal state of rest would not be eternal 
after all. In the second place, the rule that everything at rest must have been set 
at rest by something else is most strange and gratuitous. What grounds can there 
be for supposing that every state of rest is produced and has a cause? The like- 
lihood — and this addresses only the second of the two puzzling points — is that 
Aristotle was arguing not in the abstract but ad hominem; a few lines earlier he 
had been discussing Empcdoclcs, 69 and in Empedoclcs' system the world in its 
state of rest is indeed set at rest through a prior process. At any rate, the contention 
that a prccxistcnt state of rest would have had to be brought about by a cause 
docs not, as far as I could discover, recur in the Middle Ages. 

(P) Aristotle offered a second reason for ruling out alternative (ii). He submits 
that the world's beginning to move after having been stationary allows only a 
single interpretation. The relationship between the world and whatever causes its 
motion would previously not have permitted motion, whereupon the relationship 
would have changed, the motion of the world being an outcome of the change. 
But if that had happened, the supposed first motion would have been preceded 
by another motion, to wit, the change in relationship between what produces and 
what undergoes motion. The thesis that the body of the universe existed in a state 
of rest before undergoing an absolutely first motion is seen once more to imply 
a prior motion and to embody a self-contradiction. 70 

In sum: If motion should have had a beginning and the world is (i) assumed 
to be generated, its generation would have constituted a motion prior to the 
supposed absolutely first motion. If motion should have had a beginning and the 
world is (ii) assumed to be eternal — if the world is assumed to have existed in 
an eternal state of rest before starting to move — then (a) the process whereby 
the world was set in its state of rest would have constituted a motion prior to the 
supposed absolutely first motion. Moreover, (B) the change in relationship which 



* Physics VIII, 1. 251a, 20-28. 
"ibid., 250b, 26-27. 



"Ibid.. 251b. 1-10. 



( 

{ 
( 

( 20 Proofs of Eternity from Nature of World 

( initiated the first motion — that is, the change in the relationship between the 

( - world and the cause producing its motion — would also have constituted a motion 

before the supposed absolutely first motion. Having exposed the inadmissibility 

' of the two possible ways of construing an absolutely first motion, Aristotle con- 

cluded that motion cannot have had a beginning but must, together with a world 

' capable of undergoing motion, be eternal. 

( In the Middle Ages, Aristotle's complex argumentation was sometimes advanced 

( with both alternatives carefully laid out and shown to be untenable. In addition, 

his reason for ruling out alternative (i) was sometimes advanced as an independent 
( proof either for the eternity of motion or for the eternity of the celestial spheres. 

( Instances of the complete argument 

( In one of a collection of Arabic works attributed to Alexander of Aphrodisias, 

the conceivable ways of construing an absolutely first motion arc spelled out as 
C Aristotle had spelled them out: The object to undergo the first motion — that is, 

^ the physical universe — would either (i) have to come into existence, or (ii) be 

eternal. As for alternative (i), the world's "coming into existence would clearly 
( precede [its existing and moving]. . . . Coming into existence occurs, however, 

^ only through 71 motion, and consequently motion would have existed before motion 

existed." That is to say, 72 the body undergoing the first motion would have come 
f into existence through a motion preceding what was, by hypothesis, its first 

motion. The supposed absolutely first motion would have been preceded by 

another motion and would not in fact be first. 
( As for alternative (ii), it is ruled out by two considerations. The first is not the 

f one offered by Aristotle — the strange notion that if the world had been at rest 

some factor would have had to produce its rest. Instead, a substitution is made 
( and the rhetorical question is posed: (a) "How might anyone explain the world's 

starting to move now, after having been at rest for an infinite time?" 73 That 
^ question reflects a separate, widely utilized proof of the eternity of the world, 

( which is to be taken up in the next chapter. 74 The second consideration advanced 

by the text attributed to Alexander is a slight expansion 75 of Aristotle's second 
I* reason for the inadmissibility of alternative (ii). Here the argument is: (3) If the 

if world were stationary before beginning to move, "either the factor producing 

motion did not exist previously ... or else the factor producing motion did not 

stand in the requisite relation to the object that was to undergo motion." But to 
( assume either that the factor producing motion came into existence or that the 

( 

I "Aristotle reasoned that coming into existence is a motion. See Averrocs' version of the argu- 

i * 

| ment. immediately below. 

]( 72 See previous note. 

"Strictly speaking, for an Aristotelian there would have been no time without motion. 

i 1 ' 74 Below, p. 52. 

; "Aristotle did not mention the possibility that the cause producing motion was not yet in existence. 



,( 



Proofs of Eternity from Nature of World 



21 



relationship between it and the object undergoing motion changed would be to 
assume a motion prior to the supposed first motion. The assumption that the 
world was eternally at rest and thereupon began to move thus entails, once more, 
a motion prior to the first motion, which is a self-contradiction. Since the possible 
ways of conceiving a first motion are untenable, motion, the conclusion is drawn, 
must be eternal. 76 

Averroes also offers a proof of the eternity of motion which repeats Aristotle's 
distinction of the two conceivable ways for motion to have begun. Should it be 
assumed, Averroes writes, that (i) "the object to undergo the first motion came 
into existence," a prior motion would thereby be implied. In specifying what the 
implied prior motion would be, Averroes fills in a gap left by Aristotle. Aristotle 
had failed to explain why the coming into existence of the object to undergo 
motion — that is, the creation of the world — might not itself be the very first 
motion ever to have occurred; 77 the deficiency is now made up by Averroes with 
the aid of a tenet of Aristotelian physics. Four genera of motion or change are 
recognized in Aristotelian physics, and the most primary of the four is motion 
in place, a motion in place ultimately lying behind all incidents of motion or 
change in the other genera. 78 Averroes accordingly completes his elimination of 
alternative (i) by arguing that the "motion of coming into existence cannot be 
the first [motion ever to have occurred), since coming into existence is always 
dependent upon a [prior] motion in place." 79 To assume, therefore, cither that 
the physical universe came into existence and then performed the first motion to 
have occurred, or alternatively, to assume that the coming into existence of the 
physical universe was itself the first motion ever to occur, is self-contradictory. 
In either case, "the motion that was by supposition first would not in fact be 
first," since a prior motion in place would perforce be implied. 

As for the assumption that (ii) "the object to undergo motion was at rest for 
an eternity and thereupon moved," such an assumption would also imply "a 
motion prior to the supposed first motion. ... For if what produces motion and 
what is to undergo motion exist eternally . . . [and yet motion occurs only at a 
certain moment], some further factor must come into existence which induced 
the cause producing motion to produce motion and the object undergoing motion 
to undergo motion, after not having done so. But what comes into existence 80 is 



"'Alexander of Aphrodisias (?), MabadV at-Kull. in Aristu ind al- 'Arab, cd. A. Badawi (Cairo, 
1947), p. 263. 

77 Sec above, p. 18. 

'"Physics VIII, 7, and cf. below, p. 241. 

79 An added point made by Averroes is that when motion is taken in a narrow sense, the process 
of coming into existence is, although a form of change, not a form of motion. Aristotle is not consistent 
on the question whether coming into existence should, or should not, be characterized as motion. 
Sec references to Zcllcr and Ross, above, n. 62. 

""One would expect Averroes to have written: "the process of coming into existence is a motion. . . .* 



( 
( 

( 

22 Proofs of Eternity from Nature of World 

( either a motion or the result of a motion." 81 Consequently, the coming into 

existence of the factor inducing motion would constitute, or at least entail, a prior 
motion; and the supposedly first motion again turns out to have been preceded 

/ by another. The two possible ways of construing an absolute beginning of motion 

having been ruled out, motion — and hence something capable of undergoing 

( motion — must be eternal. 82 

( Versions of the complete Aristotelian proof for the eternity of motion very 

similar to Averroes' version appear in Albcrtus Magnus, 83 who could have bor- 
( rowed the proof from Averroes; in Gersonidcs, 84 who undoubtedly did; in Cres- 

!- cas, 85 who drew from either Averroes or Gersonidcs; and in Abravanel, 86 who 

copies from Gersonidcs. 

( 

Instances where the argument eliminating alternative (i) 
1 is offered as an independent proof 



( Side by side with the composite argument just examined, Averroes offers a 

proof from motion which has the specific aim of establishing the eternity of the 
heavens. His proof consists in a slight elaboration of the reasoning whereby 
( Aristotle had eliminated alternative (i) in the complete, composite argument. 

In Aristotelian physics, as already mentioned, motion in place is the primary 
' genus of motion and change, every incident of motion and change in the other 

( genera being ultimately traceable to a motion in place. Averroes now reasons: 

Motion in place is responsible for the other kinds of motion and change. Whcn- 
( ever motion and change occur in the sublunar world, what is ultimately rcspon- 

( sible is the circular motion in place of the celestial spheres. Should the celestial 

spheres themselves have come into existence, there must exist a body "prior to 
( the celestial spheres" which "undergoes motion in place . . . and is thereby 

( responsible for the spheres' coming into existence." Should the body that brought 

about the existence of the celestial spheres have itself come into existence, its 
C coming into existence must be due to the motion in place of still another body, 



( prior to it. Since an infinite series of these bodies is impossible, 87 the series must 

^ "end at [a class ofj eternal bodies undergoing a motion in place precisely like 

( 

8l Averroes adds the words "the result of a motion" because he understands that the process of 
( coming into existence is not strictly a motion. Sec n. 79. 

s2 Middle Commentary on Physics, VIII, ii, 2, taken together with Epitome of Physics, pp. 108- 
C 109. 

( "Commentary on II Sentences, in Opera Omnia, ed. A. Borgnet. Vol. XXVII (Paris, 1894), d. 

1 1, B, art. 10. 

( M Milhamot ha-Shem. VI, i, 3, pp. 299-300. 

85 Or ha-Shem, III, i, 1. A few lines have fallen out in the printed editions and I have used one 
I of the manuscripts. 

"'Mifalot, V, 1(2). 

( R7 Both because of the impossibility of an infinite number of bodies (cf. below, p. 243) and 

because of the impossibility of an inlinitc regress of causes (cf. below, p. 241). 

( 



Proofs of Eternity from Nature of World 



23 



[that of | the celestial spheres." Consequently either the celestial spheres are eter- 
nal or there exists something prior to them and precisely like them which is 
eternal. 88 

A proof of the eternity of the celestial spheres framed in the same pattern is 
recorded by Gersonidcs, 89 who undoubtedly took it from Averroes; by Crescas, 90 
whose source was either Averroes or Gersonidcs, and who condenses the argu- 
ment considerably; and by Abravanel, 91 who copies the argument from Crescas. 

A somewhat similar argument is recorded by Maimonides. Like the original 
proof of Aristotle, Maimonides' version undertakes to demonstrate the eternity 
of motion; but his formulation is plainly designed to simplify Aristotle's, where 
separate treatment had been given to the alternative that the world was created 
and the alternative that the world is eternal. Ignoring the distinction between the 
two alternatives, Maimonides' version reads: "Whatever comes into existence is 
preceded by a motion, to wit, its passing [from potentiality] to actuality and its 
coming into existence after not having existed." Therefore, "if motion itself should 
have come into existence ... [a prior] motion would already have had to exist, 
namely the motion through which motion supposedly came into existence." The 
very coming into existence of motion would, in other words, itself have been a 
motion. Should, moreover, the motion through which motion came into existence 
likewise be assumed to have a beginning, its coming into existence too would 
have been a motion. The assumption of an absolute beginning of motion thus 
entails the assumption of "an infinite regress" of motions. The assumption is 
hence contradictory and self-destructive, and motion taken collectively must be 
eternal. 92 Virtually the same argument is also recorded by Bonaventurc. 93 

Moses Narboni subsequently discovered a flaw in the reasoning. Aristotle's 
Physics had established the plausible proposition that "there cannot be motion of 
motion, or a coming into existence of coming into existence, or, in general, 
change of change." 94 What the proposition means is that the beginning of a given 
change is not a change distinct from, and prior to, the given change. Accordingly, 
Narboni insists, no true philosopher of the Aristotelian school would maintain 
that motion comes about through a process which is itself a motion; and the 
argument recorded by Maimonides is neither cogent nor genuinely Aristotelian. 95 
In its place, Narboni offers his own adaptation of Aristotle's original composite 
proof of the eternity of motion. 96 



"Middle Commentary on Physics. VIII, ii, 2. 
'"Milhamot ha-Shem. VI, i, 3, pp. 298-299. 

m Or ha-Shem. Ill, i, 1; the printed texts again have to be corrected with the aid of the manu- 
scripts. Crescas omits explicit mention of the celestial spheres, possibly because he was not certain 
of their existence. 

"'Mifalot. V, 1(2). 92 Guide, II. 14(1). 

'"Commentary on II Sentences, d. I , p. I , a. 1 , q. 2(2). 
'"Physics V, 2, 225b. 15-16. 

'"Commentary on (litide (Vienna, 1853), II. 14, p. 30a. '"'Ibid., 3()b-3la. 



24 



Proofs of Eternity from Nature of World 



Another argument of the type we are considering appears in Aaron ben Elijah. 
Aaron's version opens with the statement that "every change is preceded by a 
motion," but he furnishes no explanation as to why the statement is hue or even 
what exactly is meant by change. If the motion preceding a given change is 
preceded by another motion, and the latter by still another, an infinite regress 
ensues, which is absurd. As motions are traced back, therefore, a motion must 
eventually be reached which never underwent change, which is ever constant, 
and hence is eternal. 97 The conclusion is stronger than the conclusion of the 
argument recorded by Maimonides. Maimonides' version established that motion 
taken collectively and generically must be eternal; Aaron's version establishes 
that there must exist a single specific motion — the motion of the celestial spheres — 
which is eternal. 

Brief arguments for the eternity of motion also appear in Albertus Magnus, 98 
Aquinas., 99 and Leone Ebreo. 100 

(dj Arguments from the nature of time 

After Aristotle presented his proof for the eternity of motion, he added that 
eternity is implied by the nature of time. He brought forward two considerations. 

He argued (i) that time must be eternal because there can be no "before and 
after without time." That is to say, should time be assumed to have a beginning, 
what was before time could still legitimately be spoken of. The term before has, 
however, a temporal connotation, signifying prior in time, and therefore everyone 
who assumes a beginning of time inescapably finds himself referring to prior 
time. An absolute beginning of time is consequently impossible, and time must 
be eternal. Since time must be eternal, and since time goes hand in hand with 
motion, 101 there being no "time without motion, . . . motion too must be eternal." 
And if motion is eternal, something undergoing motion must have always existed. 
The impossibility of an absolutely first time entails, then, the eternity of time as 
well as the eternity of motion and some sort of physical world. 102 

Aristotle further argued (ii) that the concept of time and the concept of the 
moment are interrelated: "Time can neither be, nor be thought of, apart from the 
moment." But the nature of the moment is to divide past from future; for the 
moment serves as "a beginning of the future and an end of the past." The assump- 
tion of an absolutely first moment would consequently carry with it the impli- 
cation of a period of time which is terminated by, and prior to, that first moment; 



91 'Es Hayyim, chap. 6. 
"'Physics, VIII, tr. 1, chap. 11. 
"Summa Theologiae, I, 46, 1, obj. 5. 
loo Dialoghi d'Amore, p. 238. 

101 Physics V, 1 1 , 220a, 24-25: "Time is the number of motion in respect of before and after." 
m2 Physics VIII, 1, 251b, 10-13. 



Proofs of Eternity from Nature of World 



25 



and the prior time would itself contain moments. The assumption of an absolutely 
first moment is thus self-contradictory. Inasmuch as a first moment is impossible, 
lime together wilh motion ami a world undergoing motion must be eternal.'" 

These Aristotelian grounds for the eternity of time might strike us as highly 
dubious, since they rest not so much on the nature of time as on the idiosyncracics 
of human discourse and on the Aristotelian definitions of time and the moment. 104 
Nevertheless, the arguments were taken seriously in the Middle Ages. 

(i) A number of philosophers cite the argument that if time should be assumed 
to have a beginning, what was before time could still legitimately be spoken of. 
Frequently the reasoning is that, at the least, the "nonexistence ( 'adam) of time," 
could be described as preceding time. But if anything, even nonexistence, can 
be spoken of as before the assumed first time, time already existed. An absolutely 
first time is therefore impossible; and time, as well as motion, and a world, are 
eternal. The argument as thus formulated is to be found in Avicenna, 105 Ibn 
Tufayl, 106 Averroes, 107 Fakhr al-DIn al-RazI, 108 Amid!, 109 Aquinas, 110 Gerson- 
ides," 1 Iji," 2 Crescas, 113 and Abravanel." 4 

A minor variation consists in the combination of the argument from time with 
the argument from the concept of possibility, examined earlier. 1 15 The reasoning 
now is: To assume that the world was created would be to assume that the world 
had the possibility of being created before actually being created. But to speak 
of a possibility before creation is to imply a time before creation; and to imply 
time before creation is to imply that motion and a world already existed before 
creation. The assumption of the creation of a world where no world existed before 
is therefore untenable. This version of the argument, like the previous version, 

""Ibid., 19-28. 

l04 Cf. G. Hourani, "The Dialogue between al-Ghazall and the Philosophers on the Origin of the 
World," Muslim World. XLVII1 (1958), 190. 
""Najat. p. 117. 

'""//u.vv ben Yaqdhan. ed. and trans. L. Gauthicr (Beirut, 1936), Arabic text, p. 81; I'rcnch 
translation, p. 62; Knglish translation; Hayy Ibn Yaqzan. trans. L. Goodman (New York. 1972), with 
pagination of the Arabic indicated. Of the writers cited here Ibn Tufayl is the only one who does not 
reason specifically that it is the nonexistence of time which would precede time. 

[ul Long Commentary on Physics, VIII, comm. 10; Epitome of Metaphysics, cd. and Spanish trans. 
C. Quints Rodriguez (Madrid, 1919), IV, §4; German translation: Die Epitome dcr Metaphysik des 
Averroes. trans. S. van den Bergh (Leiden, 1924), p. 106. K. al-Kashf. cd. M. Mueller (Munich, 
1859), p. 34; German translation: Philosophic und Theologie von Averroes, trans. M. Mueller (Mun- 
ich, 1875), with pagination of the Arabic indicated. 

,m Muhassal, p. 91; K. al-Arba'in, p. 50. 

""Gliaya al-Maram, p. 266. 

uo Summa contra Gentiles, II, 33(6). 

'"Milhamot ha-Shem. VI, i, 3, p. 298. 

,,2 Mawaqif, VII, p. 228. 

"'Or ha-Shem, III. i, 1. 

" 4 Mifalot, V. 1(3). 

'"Above, p. 16. 



< 



26 Proofs of Eternity from Nature of World 

is found in Avicenna 1 16 and probably originated with him. It subsequently appears 
in Ghazali's critique of Avicenna's philosophy," 7 in ShahrastanI, who cites it in 
the name of Avicenna, 118 in Gersonides, 1 19 in Aaron ben Elijah, 120 and in Crescas. 121 
Another variation insists that the creator would surely have had to exist before 
any assumed creation of the world. The assumed beginning of time would for 
that reason be preceded by another time, and time would not after all have an 
absolute beginning. The philosopher responsible seems again to have been Avi- 
cenna. 122 The argument reappears in Ghazali's critique of Avicenna's philosophy 123 
and in Aquinas. 124 

Yet a further variation plays on the term when rather than the term before. The 
reasoning is that should a beginning of time be assumed, the period when there 
was no time could nonetheless be spoken of; and a time prior to the assumed first 
time would be acknowledged. The contention appears in Proclus, 125 in the medi- 
eval Arabic translation of Proclus' arguments for eternity, 126 in Shahrastanl's 
paraphrase of Proclus' arguments, 127 and in the Arabic corpus attributed to Alex- 
ander of Aphrodisias. 128 

(ii) A number of philosophers also cite Aristotle's second consideration, the 
argument that since the moment by its nature divides the past from the future, 
every moment is preceded by other moments, and an absolutely first moment is 
impossible. That line of reasoning is advanced by the Arabic corpus attributed 
to Alexander 129 and by Averroes. 130 It is recorded by Albertus Magnus, 131 Bon- 
aventure, 132 Aquinas, 133 Gersonides, 134 Crescas, 135 Abravanel, 136 and Leone 
Ebreo. 137 

" 6 Sce the passage from an unpublished work of Avicenna's quoted by S. Pines, "An Arabic 
Summary of a Lost Work of John Philoponus," Israel Oriental Studies, II (1972), 350. 
'"Tahdfut al-Falusifa, I, §81; English translation, p. 57. 
" % K. Nihaya al-Iqdam, cd. A. Guillaunic (Oxford and London, 1934), p. 33. 
,t9 Millmmot ha-Shem, VI, i, 3, p. 299. 

na, lisHayyim, chap. 6. '"Or ha-Shem, III, i, 1 . 

'"Shifa' : llahiyat. p. 379. 

' 21 Tahafut al-Falasifa, I, §§57, 70; English translation, pp. 37, 48. 
n *Summa Theologiae, 1, 46, 1, obj. 8; De Potentia, q. 3, art. 17(20). 

'"Quoted by John Philoponus, De Aeternitate Mundi Contra Proclum, cd. H. Rabe (Leipzig, 
1899), p. 103. See Philoponus' exposition, ibid., p. 104. 

'•"Published by A. Badawi in Neoplatonici apud Arabes (Cairo, 1955), p. 38. 
127 K. al-Milal wa-l-Nihal, p. 339. 

' n MabadV al-Kull, p. 264. I29 lbid. 

""Epitome of Physics, p. 1 11; Middle Commentary on Physics, VIII, ii, 3; Long Commentary on 
Physics, VIII, comm. 11; Epitome of Metaphysics, IV, §4. 
131 Commentary on II Sentences, d. 1, B, art. 10. 
'^Commentary on II Sentences, d. 1, p. 1, a. 1, q. 2. 
,3} Summa Theologiae, I, 46, 1, obj. 7; Summa contra Gentiles, II, 33(5). 
,M Milhamot ha-Shem, VI, i, 3, p. 298. 
'"Or ha-Shem, III, i, 1. 

,3S Mif'alot, V, 1(3). '"Dialoghi d'Amore, p. 238. 



Proofs of Eternity from Nature of World 



27 



(Hi) In addition to Aristotle's two considerations, the following line of thought 
is to be found: Everything that comes into existence, comes into existence in 
time. If time itself came into existence, it too would have come into existence in 
time. Time would therefore extend back beyond the assumed first time; and the 
assumption that time had an absolute beginning turns out, once more, to be self- 
contradictory. The argument is recorded by 'Abd al-Jabbar, 138 Bonavcnturc, 139 
Gersonides, 140 and Abravanel. 141 

(e) The argument from the vacuum 

Aristotle had established that a vacuum, completely empty space, cannot pos- 
sibly exist, and the impossibility of a vacuum became one of the principles of 
his physics. In one passage he utilized the principle in constructing a proof of 
the eternity of matter. 

He reasoned: Material objects can come into existence only in place. On the 
hypothesis of "absolute" generation, that is, the hypothesis that something came 
into existence from nothing, "the place to be occupied by what comes irtto exis- 
tence would previously have been occupied by a vacuum, inasmuch as no body 
existed." But a vacuum is impossible. Consequently, the generation of something 
from nothing is impossible, and matter must be eternal. 142 In the Middle Ages 
the argument from the vacuum is hinted at by Saadia 143 and is given by Aver- 
roes, 144 Aquinas, 145 Gersonides, 146 Crescas, 147 and Abravanel. 148 



"*K. al-MajimV Ji al-Muhil bi-l-Taklif, cd. J. Houbcn (Beirut. 1965). pp. 65-66. 
""Commentary on II Semem es, d. 1 . p. 1 , a. 1 . q. 2. 

'"'Milhamol ha-Shem, VI, i, 3, p. 298; 20, p. 381. Gersonides seems to attribute Ihc argument 
to Aristotle. 1 was not able to lintl it in Aristotle; but (he source could be Averroes' Middle Commentary 
on Physics, VIII, ii, 3, where Averroes writes: "What is generated and destroyed undergoes those 
processes in a moment; and the moment is the beginning of the future and the end of the past." 
Averroes probably means merely that if time began, its initial terminus would be a moment, that 
being the sense of Aristotle's Physics VIII, 1, 251b, 23. which he is paraphrasing. But Averroes 
could also be taken as meaning thai if time underwent the process of generation, it would have been 
generated in a moment, hence in time. 

""Mifatol, V, 1(3). 

' 42 Physics IV, 6-9; De Caelo III, 2, 301b, 31 ff. 
I4, A'. al-Amdnat, p. 71. 

' 44 Long Commentary on De Caelo, III, comm. 29; A'. al-Kashf, p. 34. 
,4 ' Summa Theologiae, I, 46. 1. obj. 4. 

'"'Milhamol ha-Shem, VI. i, 3, p. 301 (bottom): on VI, i, 17, p. 364. Gersonides accepts the 
argument. 

'"Or ha-Shem, 111, i, 1. 
'■'"Miftdot, IV. 3(4). 



28 



Proofs of Eternity from Nature of World 



if) The argument from the nature of the celestial spheres 

In De Caelo I, Aristotle set forth an argument for the eternity of the celestial 
spheres which rests on one of the fundamentals of his physics, his analysis of 
the process of generation. 

Aristotle understood the process of generation to consist in something's losing 
its previous character and adopting the contrary character. When a substance is 
generated, 149 a portion of matter loses its previous character, which was the 
absence, or "privation," of the form being acquired, and it adopts the contrary 
character, which is the new form. 150 The process of destruction likewise is a 
passage from one contrary to the other: In the destruction of a substance, the 
matter loses its form and is left with the absence or privation of the form, which, 
again, is the contrary of its previous character. The two processes differ in that 
the outcome is positive in the case of generation and negative in the case of 
destruction; generation begins with, whereas destruction ends with, the absence 
of a given form. 151 It follows from Aristotle's analysis that the process of gen- 
eration as well as the process of destruction can occur only where a substratum 
is amenable to contraries. For without a substratum amenable to contraries, no 
substance can be generated through the acquisition of a new form nor destroyed 
through the loss of a present form. 

Aristotle further explained that the nature of a thing expresses itself in the 
thing's motion, and contrary natures express themselves in contrary motions. Yet 
he demonstrated, or thought he demonstrated, that one type of motion has no 
contrary: "No motion is contrary to motion in a circle." 152 Now the celestial 
spheres do, by their nature, undergo circular motion. Since the motion of the 
celestial spheres has no contrary, their nature has no contrary; and whatever 
substratum the spheres have, 153 Aristotle inferred, must be of a type not amenable 
to contraries. The substratum of the spheres can, consequently, never have passed 
from a condition wherein it did not have its present nature and form to the contrary 
condition wherein it does, and the spheres can never have undergone the process 
of generation. They, together with the sublunar region whose existence is implied 
by theirs, must have always existed. 154 

'""Aristotle's analysis also covers cases where a substance lacking a certain characteristic acquires 
the characteristic, as when an unmusical man becomes a musical man. 

' S0 Physics I, 7. Either one of two things can be thought of as having come into existence: the 
abiding material substratum together with the contrary character adopted by it; or alternatively, and 
in a stricter sense, only the new character adopted by the underlying substratum. 

151 Cf. Physics V, 1. 

" 2 De Caelo 1, 4. 

'"Different positions were taken in the Middle Ages on the question whether the celestial spheres 
have a material substratum analogous to the substratum of objects in the lower, sublunar world. Sec 
H. Wolfson, Crescas' Critique of Arislolle (Cambridge, Mass., 1929), pp. 594-598. 

>54 De Caelo I, 3, 270a, 12-22. Aristotle more or less takes for granted that the existence of the 
sublunar region is entailed by the existence of the celestial region. See De Caelo 1, 8; Zcller, Die 
Philosophic der Griechen, II, 2, pp. 432-433. 



Proofs of Eternity from Nature of World 



29 



In the Middle Ages, the foregoing argument, usually much abbreviated, is 
found in Proclus, 155 Aviccnna (7), 156 Avcrrocs, 157 Aquinas, 158 Gcrsonidcs, 159 
Aaron ben Elijah, 160 Crescas,"' 1 and Abravancl. 162 

A slightly roundabout variation appears as well. It starts not with the process 
of generation but with the process of destruction; and it combines Aristotle's 
analysis of the process of destruction with the principle — also demonstrated by 
Aristotle in the De Caelo 163 — that what is not subject to destruction is not subject 
to generation. The reasoning is: Since the celestial spheres undergo only circular 
motion, their motion has no contrary and hence their nature has no contrary. The 
substratum of the spheres is therefore not amenable to contraries and can never 
pass from the condition wherein it does have its present nature and form to the 
contrary condition wherein it no longer will; and the spheres are not subject to 
destruction. But what is not subject to destruction is not subject to generation. 
The spheres, together with the sublunar region, 164 must accordingly have existed 
forever. This variation of the argument is recorded by ShahrastanI, 165 Maimon- 
ides, 166 Albertus Magnus, 167 Aquinas, 168 Gersonidcs, 169 Crescas, 170 and Leone 
Ebreo. 171 

Another variation is alluded to by Maimonides, 172 and is reported by Aqui- 
nas, 173 Aaron ben Elijah, 174 and Leone Ebreo. 175 It runs simply, and perhaps 
speciously: The celestial spheres cannot have a "beginning" because the shape 
and motion of the spheres arc circular, and the circle has no beginning. 

The arguments from (a) the nature of matter, (b) the concept of possibility, 
and (e) the vacuum, would establish the eternity of matter. They disprove creation 

"'Quoted by I'hiloponus, De Aeternitate. p. 478. 

' 56 De Caelo, chap. 4, in Opera (Venice, 1508). This seems to be a paraphrase of Themistius' De 
Caelo, and not a genuine work of Avicenna's. See M. Alonso, "Hunayn Traducido al Latin," Al- 
Andalus, XVI (1951), 37-47. 

^Epitome of De Caelo, in Rasa il Ihn Rushd (Hyderabad, 1947), p. 11; Middle Commentary on 
De Caelo (Vatican Library, Hebrew MS. Urb. 40), I, vi; Long Commentary on De Caelo. 1, comm. 
20. 

"'Summa contra Gentiles, II, 33(1). Cf. Summa Theologiae, I, 46, 1, obj. 3. 
]59 Milhamot ha-Shem, VI, i, 3, p. 300. 
>Wi Es Havyim, chap. 6. 

'"Or ha-Shem, III, i. 1. ,62 Mifalot, V, 1(4). 

,6 >De Caelo I, 12. Cf. below, pp. 91, 320. l64 Cf. above, n. 154. 

I6, A'. al-Milal wa-l-Nihal, p. 340. ' 66 Guide, II, 14(3). 

'"Physics, VIII, tr. 1, chap. 11. 

"*Summa contra Gentiles, II, 33(2). Cf. Summa Theologiae, I. 46, 1 , obj. 2; De Potenlia, q. 3, 
art. 17(2). 

'""Milhamot ha-Shem, VI, i. 3, p. 300. m 0r ha-Shem, III. i, 1. 

"'Dialoghi d'Amore, pp. 237-238. 

n2 Guide, II, 17. Cf. S. Munk's translation, he Guide des Egares, Vol II (Paris, 1861), p. 135, 
n. 2. 

m De Potenlia, q. 3, art. 17 (17, 18, and replies). i74 '£j Hayyim. chap. 6. 

n5 Dialoghi d'Amore, p. 238. 



30 



Proofs of Eternity from Nature of World 



ex nihilo, but arc compatible with the creation of the present form of the world. 
The arguments from (c) the nature of motion and (d) the nature of time would 
establish that a world always existed. Whether or not they would establish spe- 
cifically that the world we know today has existed forever would depend upon 
whether or not the eternity of motion implies the eternity, specifically, of the 
celestial spheres. 176 The argument from (f) the nature of the celestial spheres, 
would establish the eternity of the physical universe as it exists today. 

3. Replies to proofs from the nature of the world. 

The arguments we have been examining all reason from the laws of nature to 
the eternity of the world, and they invite a single overall response. An adherent 
of creation could maintain that the laws of nature govern the world as it now 
exists, but need not have governed the process whereby the world would have 
come into existence. A response more or less along that line was already made 
by Philoponus when he was dealing with arguments for eternity resting on the 
rule that "something cannot come from nothing." Philoponus explained that although 
the rule is in truth inviolable within the natural realm, it does not constrain "God, 
whose essence and actuality transcend the universe," and therefore it docs not 
prove eternity. 177 In the Middle Ages, the response that the laws of nature oper- 
ative today would not have governed the process of creation appears in various 
forms, both as an overall refutation of arguments from the nature of the world 
taken collectively and also as a refutation of the individual arguments. 

Several Kalam writers describe their opponents, the advocates of eternity, as 
having proceeded from what is "present and perceivable" (slidhid), that is, from 
what can be observed in the world today, to what is "not present and perceivable" 
{gha'ib), to the conditions that would have obtained when the world came into 
existence. But the nature of what is present and perceivable, the same writers 
object, is no infallible guide to the nature of what is not. "The Negro" is some- 
times brought as an illustration. It would be not science but the height of fool- 
ishness for the African Negro to generalize from his personal experience and to 
affirm that all mankind is black; it is no less foolish to affirm of a previous state 
of the world everything that is known about the present state. This response to 
arguments for eternity is offered by 'Abd al-Jabbar, 178 Juwaynl, 179 and the Jabir 
corpus. 180 



' 'Aristotle and Averrocs were certain that the eternity of motion does imply the eternity specif- 
ically of the celestial spheres. See De Caelo I, 2, and above, n. 88. 

"'Philoponus, as cited by Simplicius, Commentary on Physics, cd. H. Dicls, Commentaria in 
Aristotelem Graeca, Vol. X (Berlin, 1895), p. 1 141. 

™Sharh al-Usul, p. 117. 

m K. al-Shamil, p. 224. Juwaynl adds a general observation: The adherents of eternity failed to 
support their analogy by showing that the perceivable and nonpcrceivable realms arc subject to the 
same rules. 

l80 Jabir ibn Hayyan, Textes Choisis, pp. 422-423. 



Proofs of Eternity from Nature of World 



31 



A similar thought is formulated by Maimonidcs much more circumstantially 
and precisely. In a comprehensive refutation of all arguments for eternity from 
the nature of the world, Maimonides writes: "Whenever something comes into 
existence after not having existed, even in instances where the matter already 
existed and merely divests itself of one form to assume another, the nature of the 
thing after it has already come into existence ... is different from its nature 
during the process of coming into existence . . . and different as well from its 
nature before it began the process, | that is to say, different from the nature of 
whatever it might have come into existence from]. For example, the nature of 
the female seed [before pregnancy] ... is different from its nature during preg- 
nancy . . . and different as well as from the nature of a complete living being 
after the living being is born." Given the differences between the three stages 
through which generated objects pass, "no inference can be drawn from the nature 
of the thing when already existent ... to its state while progressing towards 
existence, nor can an inference be drawn from the latter to the state of the thing 
before it began to move towards existence." Proponents of eternity thus commit 
a fallacy in citing "the nature of the stable, perfected, actual universe" and 
concluding therefrom that the universe must have existed forever. The adherents 
of creation, Maimonidcs continues, do not believe that the world came into 
existence under the laws of nature operative today; such, plainly could not have 
occurred. They believe that "God brought the world, in its totality, into existence 
after nonexistence," and that the state of the world "when stable and perfected, 
in no way resembles the state of the world during its coming into existence." 
Maimonidcs proceeds to show how each of the arguments from the nature of the 
world recorded by him is resolved, once the laws now governing the world are 
understood to be different from those that would have governed the world during 
the process of its coming into existence. 181 His comprehensive response to argu- 
ments from the nature of the world reappears, somewhat condensed, in Albertus 
Magnus 182 and Aaron ben Elijah. 183 

Crcscas, too, recommends it; "Maimonidcs' comprehensive response," he finds, 
is "clearly correct, and sufficient to refute the arguments" for eternity from the 
nature of the world. 184 In discussing if Crescas does not, however, employ Mai- 
monidcs' own terminology, but borrows a formulation from Gcrsonides. 185 He 
has Maimonidcs distinguish between "general coming into existence," that is, the 
coming into existence of the world as a whole, and "partial coming into exis- 
tence," the coming into existence of objects within the world. 186 The error in 

'"Guitle, II. 17. 

'"Physics, VIII, tr. 1, chap. 14; Commentary on II Sentences, d. 1, B. art. 10. 

181 -Es Hayyim, chap. 6. ""Or ha-Shem. Ill, i, 4 (beginning). 

" 5 Cf. below, p. 32 f. In Milhamoi ha-Shem, p. 306, Gcrsonides seems to attribute the formulation 
to Maimonides, but on p. 366, he quotes Maimonidcs' genuine formulation. 

""■Maimonidcs distinguished between the separate stages in the generation of an object, whereas 
Crcscas has him distinguish between the part and the whole. 



32 Proofs of Eternity from Nature of World 

arguments from the nature of the physical world is, as Crcscas explains Mai- 

monides' intent, that they draw an improper "analogy between partial coming 
{ into existence and general coming into existence." The laws governing the coming 

into existence of objects within the world were transferred by the adherents of 
1 eternity to a phenomenon of a different kind, the coming into existence of the 

( world as a whole; since the coming into existence of the world as a whole need 

not have been governed by the laws of partial coming into existence, the argu- 
ments in question are all fallacious. 187 
( Aquinas similarly offers a comprehensive refutation of arguments for eternity 

from the nature of the world and he arrives at the same result as the writers 
1 already quoted, although I could see no evidence of direct filiation. 188 The dis- 

( tinction Aquinas delineates is that between the realm of "nature," on the one 

hand, and the realm of "divine" action, on the other. In the realm of nature, there 
'< occur "change" (mutatio) and "particular" production, that is to say, the produc- 

( tion of a particular object from another particular object; whereas in the realm 

of "divine" action, there occur "creation," "simple emanation," and "universal 
' production," that is to say, the bringing forth of being when nothing at all pre- 

f viously existed. Arguments for eternity from the physical world reason from laws 

that relate to the phenomenon of particular production, but not necessarily to the 
( phenomenon of universal production or creation. As a consequence, the argu- 

~l ments are intrinsically invalid. 189 

The foregoing refutations of arguments from the nature of the world invalidate 
\ every argument of the sort, without exception. Nothing in the present state of 

I the world, the recurrent objection goes, is pertinent to the issue of creation and 

eternity, since the laws of nature operative today need not have been operative 
( during the process of creation. Gersonides takes a separate tack. He too proposes 

^ an overall refutation of arguments for eternity from the nature of the world, but 

his refutation is qualified and restricted. A certain group of arguments from the 

nature of the world is, for Gersonides, fallacious, whereas another group is valid. 
( Gersonides begins with the distinction, which in one passage he appears to 

attribute to Maimonides, 190 between "partial coming into existence" and "general 
( coming into existence." And he warns against indiscriminate "analogy" — to be 

( specific, against analogies wherein "we affirm of [general] coming into existence, 

everything that can be affirmed of the coming into existence of each single part 
f of the world." But the distinction between partial and general coming into exis- 

^ tence is found by Gersonides to be too loose for a definitive resolution of the 

issue of creation, inasmuch as some characteristics of partial coming into exis- 
( tence can properly be affirmed of general coming into existence. Only "what 

/ 

/ ™ Or ha-Shem, III, i, 2. 

188 Aquinas' refutation is very-similar to Philoponus', above, n. 177. 
( 18S 'Commentary on Physics, VIII, §§974, 987; Summa Theologiae, I, 45, 2, ad 2. 

,90 Milhamot ha-Shem, VI, i, 4, p. 306; cf. above, p. 31, and n. 185. 

/ 
( 



Proofs of Eternity from Nature of World 



33 



belongs to the world because it is in its present state (fo'ar)," he understands, 
"need not have belonged to the world before it was in that state. What, by contrast, 
belongs to the world simply because it exists, irrespective of its state, would have 
had to belong to the world even during the period of its coming into existence." 191 
The unqualified distinction between partial and general coming into existence 
must accordingly be supplemented, or replaced, by a distinction between char- 
acteristics that arc and those that are not linked to the world's existing in its 
present state. Characteristics linked to the world's existing in its present state 
would not appertain to the world's coming into existence as a whole; but char- 
acteristics of existent beings, not related to any given state of the world, would. 
Consequently, any argument for eternity which takes its departure from charac- 
teristics of the former sort is invalid, since the characteristics in question would 
have been absent — as Maimonides and other advocates of creation had insisted — 
during the world's coming into existence. But arguments from characteristics of 
the latter sort, from characteristics of existent beings which arc not related to any 
given state of the world, retain their validity; for even at the moment of its 
generation the world would have been something existent. Gersonides criticizes 
Maimondes for having failed to take account of the distinction between these two 
sorts of characteristics and for having therefore rejected, without exception, every 
inference from the state of the world when already existent to the state of the 
world during the process of its coming into existence. 192 One may wonder how 
Gersonides could recognize and identify characteristics of things which are and 
characteristics of things which are not due to the present state of the world, but 
he seems to have thought that the identification could be made intuitively. 193 It 
will appear that his qualified refutation of arguments from the nature of the world 
has the effect, as he applies it, of sanctioning arguments for the eternity of matter, 
while at the same time ruling out arguments for the eternity of the form of the 
world. 

Thus far we have seen the comprehensive refutations of arguments for eternity 
from the nature of the world. In addition to their comprehensive refutations, the 
advocates of creation offered individual refutations for each of the several arguments. 

(a) Responses to the arguments from the nature of matter 

The argument from the nature of matter, stated briefly, had been that whatever 
comes into existence does so from a preexisting substratum, or matter; matter 
too could only have come into existence from a preexisting substratum, or matter; 
consequently, to assume an absolute coming into existence of matter is self- 
contradictory. 194 



m Milhamot ha-Shem, VI, i, 4, p. 304; 17, p. 366. 
I92 lbid., 4, p. 306; 17, p. 366. 
l93 Cf. ibid., 17, pp. 364-366. 
""Above, p. 13. 



34 



Proofs of Eternity from Nature of World 



The sole grounds Aristotle provided to support the rule that whatever comes 
into existence does so from an already existing substratum were inductive; other- 
wise he treated the rule as self-evident. 195 And as has been seen, writers asso- 
ciated with the Kalam school repeatedly ascribe their adversaries' position to 
induction or analogy. 196 When Kalam writers take up the task of refuting their 
opponents, their reply consists largely in exposing the unreliability of inductions 
and analogies from what is empirically known. 

Saadia and Ibn Hazm had represented the advocates of "eternity" (dahr) as 
accepting nothing but the reports of the senses and as rejecting the creation of 
matter and the world because sense perception cannot attest to an instance of 
creation. In response, Saadia and Ibn Hazm point out that no man has ever had 
sense perception of the eternity of the world or a single one of its parts. Con- 
sistency therefore would demand that anyone who relies exclusively on analogy 
and induction from what he perceives should reject not merely the creation of 
matter and the world, but also the eternity of matter and the eternity of the world. 
Saadia goes on to list various items of nonempirical knowledge — such as mem- 
ories and inferences, including inductions themselves — which are not directly 
acquired through sense perception, yet are perforce accepted by all mankind 
including the advocates of eternity. The conclusion he and Ibn Hazm reach is 
that uncompromising empiricism is indefensible and that "analogy" from what 
is perceived through the senses cannot settle the issue of eternity and creation. 197 

'Abd al-Jabbar, Juwayni, and the Jabir corpus respond to the argument from 
the circumstance that no instance of creation ex nihilo has ever been observed 
by pressing their comprehensive refutation of arguments from what is "present 
and perceivable." The nature of the perceivable, they contend, is no reliable guide 
to the nature of what is not perceivable, and events need not always have occurred 
as they now are perceived to occur. 198 

Maimonides and advocates of creation who follow his lead refute Aristotle's 
argument from the nature of matter by applying their own comprehensive refu- 
tation of arguments from the nature of the physical world. Maimonides concedes 
that the underlying matter of the universe cannot be generated and have come 
into existence in the manner in which generated-destructible things are generated 
and come into existence; for, as Aristotle correctly held, all generated-destructible 
objects are generated from a preexisting substratum. The belief of the advocates 
of creation, Maimonides explains, is that "God created (awjada) matter from 
nothing" and that the act of creation is entirely different from the process of 
generation as it takes place within a stable world. Creation consequently does 



195 Above, p. 14. 
196 Above, p. 15. 

,97 Saadia, K. al-Amanat, I, 3, pp. 63 - 65; Ibn Hazm, K. al-Fasl ft al-Milal. I, p. 10. 
198 'Abd al-Jabbar, Sharh al-Usul, p. 117; Juwayni, K. al-Shamil, p. 224; Jabir, Textes Choisis, 
p. 422. 



Proofs of Eternity from Nature of World 



35 



not require a preexisting .substratum; thus Aristotle's argument from the nature 
of matter has no bearing on the doctrine of creation ex nihilo. 199 Albcrtus Mag- 
nus, 200 Aquinas, 201 Aaron ben Elijah, 202 Crcscas, 203 and Leone Ebrco 20 ' simi- 
larly rebut the argument from the nature of matter by distinguishing, through one 
formula or another, the phenomenon of generation within the world from the 
completely different phenomenon of the creation of the world in its entirety. 

Aaron ben Elijah appends a further consideration in an ad homincm mode. He 
notes that Aristotelian philosophy recognizes instances of things' coming into 
existence from nothing. When a new object comes into existence within the world, 
its material side is indeed drawn from already existing matter. But the form of 
the new object comes neither from the already existing matter nor from anything 
else whatsoever; the form comes from nothing. Since the Aristotelian adherents 
of eternity do not gainsay the constant coming into existence of forms from 
nothing, 205 how, Aaron marvels, can they balk at the coming into existence of 
matter from nothing'.' 206 

In contrast to the foregoing, Gcrsonidcs' overall response to arguments for 
eternity from the nature of the world 207 is formulated in a way that lets the 
argument from the nature of matter stand. Gcrsonidcs excludes any inference 
regarding the coming into existence of the world as a whole which reasons from 
characteristics tied to the present state of the world. The impossibility "that a 
body should come into existence from . . . absolutely nothing" is not, he under- 
stands, such a characteristic. It is rather a universal law of physical existence, 
"unrelated to the state in which the world exists." 208 To assume that the physical 
universe came into existence from absolutely nothing therefore embodies for 
Gcrsonidcs, as for Aristotle, a self-contradiction, and Gcrsonidcs concludes that 
matter is eternal. His considered position is that the world was created from a 
precxistent eternal matter. 209 



'""Guide, 11, 17. 

Physics, VIII, Ir. 1, chap. 14. 
2m Summa Theologiae, I, 46, 1, ad 3; Summa contra Gentiles. II, chap. 37(1). 
2 " 2, Es Hayyim. chap. 6. 

2 "'Or ha-Shem. Ill, i, 5. 2M Dialof;hi d'Amore. p. 239. 

2 "'ln Metaphysics VII, 8, 1033b, 5-6. Aristotle stated that forms are not generated; but ibid., 
15, 1039b, 26, he conceded that forms, although not generated, sometimes "are" and sometimes 
"arc not." Cf. Ross's note to 1033b, 5-6. The anomalousncss of Aristotle's position is underlined 
by Zcllcr, Die Philosophic der Griechen, II, 2, pp. 347-348, and C. Bacumker, Das Problem der 
Malerie (Mucnster, 1890), pp. 287-288. 

J0fi 7i.v Hayyim. chap. 6. The same point was made by I'hiloponus; see De Actcrniiaie. p. 351, 
and Simplicius, Commentary on Physics, p. 1 142. 

207 Above, p. 32. 

2m Milhamot ha-Shem. VI, i, 17, pp. 365-366. Abravanel rejoins that the impossibility of some- 
thing's coming into existence from absolutely nothing is in fact a characteristic tied to the present 
stale of the world; Mif'alol. IV, 3. 

''Milhamol ha-Shem. VI, i, 17, pp. 367-368. To be more precise, Gcrsonidcs' position is that 



201, 



36 



Proofs of Eternity from Nature of World 



(b) Responses to the argument from the concept of possibility 

The argument had been: Prior to something's coming into existence, there is 
a possibility of its existing. The possibility of existing must subsist somewhere; 
and it can only subsist in an already existing substratum, hence in an already 
existing matter. The assumption that matter came into existence from absolutely 
nothing consequently embodies a self-contradiction, and matter must be eter- 
nal. 210 Medieval advocates of the creation of the world developed three responses 
to the argument. 

One response consists in referring the possibility of matter's coming into exis- 
tence to the agent that produced matter. The premise is accepted according to 
which the possibility of matter's existing would have to precede the actual exis- 
tence of matter. But the prior possibility of matter's existing is not located in a 
substratum, from which matter would have come into existence. It is instead 
identified with the power of the creator to create. When the possibility of the 
existence of matter is so construed, the assumption of creation ex nihilo no longer 
contradicts itself by implying the prior existence of a substratum containing the 
possibility. 

This response to the argument from the concept of possibility is mentioned, 
but not seriously, by Avicenna, 2 " the apparent author of the argument that is 
being rebutted. It is also mentioned in passing by Ghazali 212 and is employed by 
Aquinas, 213 Aaron ben Elijah, 214 Crcscas, 215 and Abravancl. 216 The same response 
was known as well to Averroes, who attributed it to John Philoponus, and to 
Maimonides, who attributed it to the later Kalam. 217 Both Averroes and Mai- 
monides reject it because of a certain distinction that had been drawn by Aristotle. 

Aristotle had distinguished two 5vvau,eis — "powers," or "potentialities," or 
"possibilities" — in the process of change, namely the power, or possibility, of 
the agent to effect the change, and the power, or possibility, of the object to 
undergo the change. 218 Accordingly, Averroes and Maimonides maintain, although 
one possibility of the existence of matter may properly be identified as the agent's 



the world was created from a preexistent "body free of all form"; in the passage cited, he attempts 
to elucidate the concept. 
2,0 Abovc, pp. 16-17. 

2U K. al-lsharat wa-l-Tanbihat, ed. J. Forget (Leiden, 1892), p. 151; French translation, with 
pagination of Arabic indicated: Livre des Directives el Remarques, trans. A. Goichon (Beirut and 
Paris, 1951). 

2,2 Tahafut al-Falasifa, I, §85; English translation, p. 59. 

2 ' 3 Summa Theologiae, I, 46, 1, ad 1; Summa contra Gentiles, II, chap. 37. 

214 'Es Hayyim, chap. 6. 

2 "Or ha-Shem, III, i, 5. 

2,(, Mifalot, IV, 3. 

217 Avcrrocs, Long Commentary on Metaphysics, XII, comm. 18; Averroes writes that he drew his 
information from Alfarabi's work On Changeable Beings. Maimonides, Guide, II, 14(4). 
"'Metaphysics V, 12, 1019a, 15-22; cf. ibid., IX, 1. 



Proofs of Eternity from Nature of World 



37 



power to create matter, there remains the question of the other possibility of the 
existence of matter, the possibility from the side of some object that is to undergo 
the process of becoming matter. Until the latter possibility is explained or explained 
away, the argument that the prior possibility of the existence of matter would 
have to be located in an already existing substratum will not have been answered. 219 
Perhaps because of the point raised by Averroes and Maimonides, some writers — 
Aquinas, Aaron ben Elijah, and Abravancl — do not utilize the first response to 
the argument from the concept of possibility by itself, but always buttress it with 
additional considerations. 220 

The second response to the argument from the concept of possibility seems to 
have originated in Ghazali. Ghazali construes the "possibility" of something's 
coming into existence — along with the impossibility, and the necessity, of some- 
thing's coming into existence — as nothing but an "intellectual judgment," a judg- 
ment on the part of the intellect that the thing may — or that it cannot, or that it 
must — exist. Since possibility, impossibility, and necessity, have no objective 
existence in the external world, they do "not require anything existent" to serve 
as their substratum. And since the possibility of matter's existing docs not require 
a substratum, the argument from the concept of possibility has no validity. 221 

Following Ghazali, ShahrastanI construes the possibility of matter's coming 
into existence as a "mental supposition" (taqdir); 222 Fakhr al-DIn al-RazI denies 
that the possibility of matter's coming into existence is an "existent attribute"; 223 
Amid! writes that it is not "a real essence." 224 Aquinas offers an interpretation 
in the same vein, and even finds support in Aristotle. In the course of analyzing 
the divers meanings of the term "possible" (8uvaTov), Aristotle had isolated 
what we should call the logically possible. And, he had stated, the possible in 
its logical sense carries no implication of a power — or potentiality, or possibility 
(Svvaixis)— in either an agent producing change or an object undergoing change. 225 
Aquinas calls attention to Aristotle's statement; and he explains that the coming 
into existence of matter and of the world was possible "in the way a thing is said 
to be absolutely possible, that is, not by virtue of any potentiality, but solely 
from the relation of the terms, which arc compatible with one another." 22 '' The 
possibility of existence preceding the actual existence of matter was thus not 
anything with an external existence of its own. A possibility of existence preceded 



2>v Averroes, Tahufiit al-Tahdfut, I, pp. 100-101; Maimonides, Guide, II, 14(4). 
220 Sce below, p. 38. 

22l Tahdful al-Falasifa, I, §87; English translation, p. 60. 
22Z K. Niluiya al-Iqddm, p. 34. 
2 "Muhassal, p. 91, K. al-Arba'in, p. 51. 
224 Ghaya al-Maram, p. 272. 

225 Metaphysics V, 12, 1019b, 27-29, 34-35. This is the chapter in the Metaphysics referred to 
above in n. 218. 

226 Summa Theologiae, I, 46, 1, ad 1. 



< 

I 

38 Proofs of Eternity from Nature of World 

I the actual existence of matter and of the world only inasmuch as the terms' in the 

proposition 'matter exists' or 'the world exists' were logically compatible — only 
\ inasmuch as the proposition involved no logical impossibility — before those things 

^ did exist. Aquinas combines the previous response to the argument from the 

concept of possibility with the response we are now examining. When the pos- 
( sibility of the existence of the world is said to have preceded its actual existence, 

the meaning, Aquinas writes, is either that the creator had the power to create 
1 the world, or else that the creation of the world was logically possible. 227 

( A third response to the argument from the concept of possibility consists in 

. applying the comprehensive refutation of all arguments from the nature of the 

world. The requirement that the possibility of existence must be located in a 
( substratum is held to be a characteristic of change and generation within the 

world, but not necessarily a characteristic of the creation of a world. 
Such is Maimonidcs' response. He states that what comes into existence must 
( be preceded by its possibility "only in our stable universe, where things are 

^ generated solely from something existent. When, by contrast, something is cre- 

ated ex nihilo, nothing at all existed [previously . . . which might permit the 
f thing coming into existence] to be preceded by possibility." 228 The same inter- 

pretation is given by Albertus Magnus 229 and Leone Ebreo. 230 TusT rejects the 
previous response to the argument from possibility, the response that construes 
( possibility as a "nonreal" (ghayr wujudi) attribute; for, according to TusT, the 

possibility of being generated is unquestionably a "disposition" that "requires a 
' subject" in which to inhere. In preference to the previous response, TusI offers 

{ the one now being examined. He distinguishes between the generation of things 

! within the world and creation; and he contends that creation is a completely 

j' different phenomenon, that in the case of "created things, no disposition is con- 

\( ccivablc prior to their existence." 231 Aaron ben Elijah joins the third response to 

arguments from the concept of possibility with the first. He writes: The possibility 
^ of a thing's coming into existence precedes the actual existence of the thing only 

( in the "stable, settled universe" but would not do so, in the creation of matter ex 

nihilo; moreover, whatever possibility there might be in the case of creation ex 
f nihilo is to be referred to the "agent," that is to say, the creator. 232 Abravanel 

\t similarly explains that in "first creation" as distinct from "generation," no "pos- 

| sibility in a substratum" precedes the process of coming into existence. The sole 

true possibility at that stage is the "possibility of action" on the part of God's 
If "infinite power." 233 Aquinas in one passage combines all three solutions. He 

( 227 Ibid.; De Polentia. q. 3, art. 17 (ad 10). 

/ 22s Guide, II, 17. 

229 Physics, VIII, tr. 1, chap. 14. 
/ ""Dialoghi d'Amore, p. 239. 

"'Glosses to Razl's Muhassal, p. 92. 
\ 232 'Es Hayyim, chap. 6. 

"'Mifalot. IV, 3. Sec below, n. 254. 

/ 

( 



Proofs of Eternity from Nature of World 



39 



maintains that the phenomenon of creation is radically different from the phe- 
nomenon of things' coming into existence by motion; in the former, as distinct 
from (he latter, (he prior possibility of existence is to be referred to die agent; 
alternatively, it is to be construed as nothing other than the logical compatibility 
of the terms in the proposition "matter exists." 234 

(c) Responses to arguments from the nature of motion 

Aristotle's argument from the nature of motion, it was seen, reappears in the 
Middle Ages in its original complex form as well as in a simplified form. The 
contention in both forms of the argument was that the supposedly very first motion 
would, by virtue of the laws of motion, have to be preceded by another motion, 
and hence the supposedly first motion would not in fact be first. 235 The adherents 
of creation who respond to the proof in one or the other of its forms all adduce 
the comprehensive refutation of arguments from the nature of the world. They 
distinguish between natural processes within the world and creation; and they 
explain that argumentation from the nature of motion can be valid only where 
the laws of nature arc operative but can have no bearing on God's creating the 
world ex nihilo. This response is given by John Philoponus, 230 and it is repeated 
in the Middle Ages with minor variations by Maimonidcs, 237 Albertus Mag- 
nus, 238 Bonaventure, 239 Aquinas, 240 Gersonidcs, 241 Aaron ben Elijah, 242 Cres- 
cas, 243 Abravanel, 244 and Leone Ebreo. 245 

(d) Responses to arguments from the nature of time 

The proof for eternity from the nature of time reasons in one fashion or another 
that anyone assuming an absolute beginning of time cannot avoid recognizing a 
prior time; hence time cannot have a beginning but must be eternal; time, how- 
ever, involves motion and a moving body; and time being eternal, motion and a 
world undergoing motion must also be eternal. 24 '' A response might proceed 
either by rebutting the premise that time involves motion and a moving body, or 
else by rebutting the grounds that had been advanced for the eternity of time. 
The only writer I could find taking the former course is Aaron ben Elijah. In 
reply to one of the versions of the proof, he acknowledges that time is eternal, 
but denies that time involves motion. "Time," he asserts, "is something extrinsic 

"\Summa contra Gentiles. II, 37(3). 2,5 Above, pp. 18-20. 

2 "'Quotcd by Simplicius, Commentary on Physics, pp. 1 141, 1 150. 
"''Guide. 11, 7. 

"'Commentary on II Sentences, d. 1 , B, art. 10. 
"''Commentary on II Sentences, d. I , p. 1 , a. 1 , q. 2, ad. 2. 

""Commentary on Physics, VIII, §987. Summa Theologiae, I, 46. 1, ad 5; Summa contra Gen- 
tiles. II. 36(3). 

24, Milhamot ha-Shem. VI. i, 4, p. 304; much elaborated and nuanced. ibid., 24. p. 395. 
242 Hayyim, chap. 6. 

24, Or ha-Shem. III. i, 4 (beginning). 244 Mifalol. V, 3. 

"'Dialoghi d'Amore, p. 239 . 246 Above, pp. 24- 25. 



40 



Proofs of Eternity from Nature of World 



to motion" and motion is not entailed by it. Arguments establishing the preex- 
istence of time are therefore irrelevant to the issue of creation. 247 

Most proponents of creation from Plato onwards take the other course. They 
accept the premise that time involves motion and a moving body, but maintain 
that time is not eternal, that there was no time before the world was created. 248 
These proponents of creation had to refute Aristotle's arguments for the eternity 
of time. 

(ij The first argument for the eternity of time ran: On the assumption of an 
absolute beginning of time, the period before time and when there was not yet 
time could still be spoken of; but the terms before and when imply time; time 
would thus already have existed before the assumed absolute beginning of time. 
The tenor of medieval refutations was set, or anticipated, by John Philoponus. 
Philoponus was addressing the version of the argument according to which the 
words 'when there was no time' imply that time already existed. He explains that 
with care the mischievous expression is easily avoided. A careful speaker can 
avoid mentioning the period "when there was no time" and restrict himself to 
saying, "simply, that time is not eternal." 249 But the complete solution, Philo- 
ponus holds, goes deeper. The complete solution lies in understanding that the 
argument is "sophistical," since it is concerned with "wording" rather than with 
what is "meant" thereby. Nothing can be inferred about reality from the "weak- 
ness" of human speech, and attention should always be directed to "intent" rather 
than "words." 250 The expression "when there was no time" and similar expres- 
sions consequently shed no light on the issue of creation and eternity. 

In the Middle Ages, it is repeatedly stated that language seeming to connote 
time does not imply actual time. Philoponus' distinction between intent and words 
does not explicitly appear, but another motif that may have derived from him 
does. Aristotle had affirmed the fir.iteness of space side by side with the infinitc- 
ness of time, and proponents of creation pounce on the apparent inconsistency. 
They draw an analogy between temporal and spatial extension and contend that 
just as space is universally acknowledged to have a terminus beyond which there 
is no space, so too may time have a beginning before which there was no time. 
For example, the following critique of the Aristotelian doctrine of eternity is 
reported in the name of "Yahya," that is to say, Yahya ibn 'Ad! or, possibly, John 
Philoponus: 251 The proposition that time began at a certain moment with no time 



247 'Es Hayyim, chap. 6. Aaron is answering the argument that at the very least the possibility of 
the existence of the world would be present before the world came into existence, and hence time 
would already exist before the supposed beginning of time. 

248 Sec Plato, Timaeus, 38; Philo, De Opificio Mundi, vii, 26; Augustine, City of God, XI, 6; 
Confessions, XI, 30; H. Wolfson, Philo, Vol. I (Cambridge, Mass., 1948), p. 311. 

lw De Aeternitate, p. 105. 

250 Ibid., pp. 104, 116. 

2 "ln Arabic, Philoponus is called Yahya al-Nahwi. 



Proofs of Eternity from Nature of World 



41 



preceding is neither more nor less admissible than the proposition that space 
terminates at a certain point with nothing beyond. It is a familiar tenet of Aris- 
totelian physics that the world is finite and that "no extension whatsoever," neither 
"plenum nor vacuum," exists outside the world. By what right then do the Aris- 
totelians reject a parallel theory in regard to time? By what right do they brand 
as "inconceivable, the thesis that there was no time" before creation, that time 
extends back to a certain moment with absolutely no time preceding? 252 

Ghazali weaves together two motifs, the contention that the vagaries of lan- 
guage shed no light on the issue of creation, and the analogy between temporal 
and spatial extension. The former motif is developed in a manner akin to his 
response to the arguments from the concept of possibility. There he contended 
that the possibility preceding the creation of the world is merely a judgment of 
the mind, a logical judgment, representing nothing in the external world; 25 '' here 
he contends that the time preceding the creation of the world can represent nothing 
in the external world because it is purely imaginary. 254 

Ghazali is answering the argument that the statement 'God exists before cre- 
ation' implies a time before the assumed creation of the world. He replies that 
the issue cannot be settled through the testimony of human language. The state- 
ment that God existed before creation can be recast in the form: "God was, 
without a world and without time, then was, with a world and time"; and the 
implication of a time before creation vanishes. 255 But in any event, even if a 
speaker should not trouble, or should be unable, to avoid language with temporal 
connotations in referring to what was before time, no actual prcexistent time is 
thereby implied. For the creation of the world is preceded by time only in our 
"imagination" ; creation is preceded by nothing more than a "supposition" (taqdir) 
of time. The situation is exactly analogous to that obtaining in spatial extension. 



2,2 Abu al-Faraj's (?) comment u> Physics VIII. 1; in Aristotle, Tub}' a (medieval Arabic translation 
of Physics together with four medieval Arabic commentaries), ed. A. Iladawi (Cairo, l%4). p. 816. 
It is not clear from the printed text whether Abu al-Faraj is quoting "Yahya" or whether the passage 
belongs not to Abu al-Faraj*s commentary but to the commentary of Yahya ibn 'Adi. 

2 "Abovc, p. 37. 

2M Thc later Kalam writers who maintained that the possibility of matter's coming into existence 
is not real (above, p. 37). and especially Shahrastani (above, n. 222). may well have been following 
Ghazali's response to the argument from time, rather than his response to the argument from the 
concept of possibility. They may have meant, in other words, not that the possibility of matter's 
coming into existence is a logical judgment, but that it is a product of the imagination. An unambig- 
uous instance where the response to the argument from time is applied to the argument from the 
concept of possibility can be found in Abravanel. After giving the reply to the argument from the 
concept of possibility quoted above, n. 233, Abravanel adds: "It also can be stated that the subject 
of this possibility is in a certain sense the human intellect. Just as a beginning of time is inconceivable 
without ... a time extending beyond . . . and prior thereto. . . . so too it is difficult to conceive of 
the actual coining into existence of an object without the notion that the possibility of the thing 
precedes its actual existence. . . . But that is only a mental precedence." 

2>i Tahafut al-Falasifa, 1, §58; English translation, p. 38. 



42 



Proofs of Eternity from Nature of World 



Nothing whatsoever lies beyond the boundary where the world and its space 
terminate, although the human imagination "balks" at the notion and mistakenly 
insists upon "supposing" an empty space beyond the world. Similarly, no time 
whatsoever existed before the world and time came into existence, although the 
human imagination again balks and insists upon "supposing" a prior time. 256 

In rebuttals of arguments for the eternity of time, the phrases imagination, or 
supposition of time become catchwords; and the analogy between the termination 
of space and the beginning of time also recurs. ShahrastanI answers arguments 
of the type we are examining by explaining that any implication of time prior to 
the world is "an imaginary supposition, like the supposition of a vacuum beyond 
the world." 257 Merely imagining that a vacuum exists beyond the world does not 
establish that a vacuum exists in actuality. By the same token, merely imagining 
that time existed prior to the world docs not establish that time truly existed. 
Maimonides, who does not include an argument from time in his formal classi- 
fication of proofs for eternity, forestalls such an argument. The statement "God 
existed before creating the world" and other phraseology that seemingly implies 
time prior to creation refer, Maimonides writes, solely to the "supposition of 
time or to the imagination of time, and not to real time." 258 Fakhr al-Din al-RazT 
maintains that "priority is not an existent attribute" 259 or a "positive attribute." 260 
Amid! characterizes the assumption of a temporal extension prior to the world as 
an "imaginative supposition." 261 Aquinas writes that the "before we speak of as 
preceding time" does not refer to time "in reality, but solely in [our] imagination." 
The situation, he adds, is analogous to "saying that there is nothing above the 
heavens." In that sentence, the term above does not have a genuine spatial ref- 
erence but merely an imagined reference, and all will acknowledge that no actual 
space above the heavens is implied. By the same token, the words "before time" 
should be acknowledged to contain no implications of an actual prior time. 262 

Gcrsonidcs' response to the argument from the nature of time is similar to the 
responses made by his predecessors, but — as Gcrsonidcs often docs when repeat- 
ing commonplaces — he appends a small twist of his own. The statement that 
time did not exist "before" the creation of the world is, he writes, analogous to 
the statement that "neither a vacuum nor plenum is to be found outside the world." 
The term before in the one instance and the term outside in the other are employed 



256 Ibid . , §§61-63; English translation, pp. 41-42, and cf. van den Bcrgh's note. 

257 A'. Nihaya al-Iqdam, p. 52. 

"'Guide, II, 13(1). 

259 Muhassal, p. 91. 

2m K. al-Arba'in, p. 53. 

2(,] Ghaya al-Maram, p. 272. 

262 Summa contra Gentiles, II, 36(6). Cf. Commentary on Physics, VIII, §990; Summa Theologiae, 
I, 46, 1, ad 8. 



Proofs of Eternity from Nature of World 



43 



equivocally and not in their ordinary sense, and hence their use cannot serve as 
grounds for inferring cither the infinity of space or the eternity of time. We have 
here, Gcrsonidcs continues, an area where human language, which is in any event 
"conventional," fails to provide the technical terminology required by philosophy 
and science; the result is that philosophers must use ordinary words in a "bor- 
rowed" sense. 263 A similar reply to the argument is made by Abravancl. 264 

(ii) Aristotle's second argument for the eternity of time was that the moment 
invariably divides past from future; to assume a first moment would therefore 
amount to assuming a preceding past time that terminates at the first moment; 
and the assumption of an absolutely first moment is self-contradictory. 265 The 
argument did not receive as much attention as the previous argument from time, 
but several responses can be cited. 

Philoponus condemns the argument as "begging the question," since in assum- 
ing that every moment divides past from future, the eternity of time is presup- 
posed from the outset. To elucidate his meaning, Philoponus draws another anal- 
ogy between space and time. Aristotle had once characterized the moment in 
time as analogous to the point on a line, 266 inasmuch as the point divides a line 
into two segments and the moment divides past time from future time. It clearly 
would, Philoponus writes, be a begging of the question to presuppose that every 
point, without exception, divides a line into two segments, to infer therefrom 
that no point can ever stand at the terminus of a line segment, and to conclude 
that every line must be infinite. It is no less begging the question to presuppose 
that every moment, without exception, divides past time from future time, to 
infer that no moment can stand at the very beginning of time, and to conclude 
that time must be eternal. 267 

Alfarabi and Averrocs knew that the argument from the nature of the moment 
might be rebutted on the grounds that the moment in time is analogous to the 
point on a line, and every line segment docs begin at a point. 268 Any rebuttal of 



2b, Milhamot ha-Shem, VI, i, 21, pp. 384-385. 

2M Mifatot. V. 3. 265 Above, p. 24. 

2 "*I'hysics IV, II, 220a. 9-10; 13. 222a. 13. 

267 Quotcd by Siniplicius, Commentary on Physics, p. 1 167. 

^Epitome of Metaphysics, IV, §4, with reference to Alfarabi; Tahiifut al-Tahafut. I, p. 77. It very 
likely was thanks to Alfarabi and Averroes that Philoponus' rebuttal reached the Middle Ages. 
Averrocs reports that Alfarabi's book On Changeable Beings, now lost, laid bare the sophism in any 
comparison between the possibility of a beginning of time at a moment and the possibility of a 
beginning of a line at a point. As far as I could see, Averroes does not mention Philoponus in this 
connection. But refuting Philoponus is known to have been one of Alfarabi"s objectives in his book 
On Changeable Beings (sec above, n. 217; below, p. 128; M. Stcinschncider. M-Farabi ( St. Peters- 
burg, I869|, pp. 120-122). It may therefore well have been just Philoponus whom Alfarabi had in 
mind in the passage referred to by Averrocs, that is to say, in the surrejoinder to the rebuttal of the 
argument from the nature of the moment. 



< 

\ 



44 Proofs of Eternity from Nature of World 

the sort is dismissed by them as a sophism, 269 but several medieval adherents of 

creation employ it nonetheless and they for their part dismiss their adversaries' 

imputation of sophistry. 
v Aquinas makes a reply to the argument from the nature of the moment which 

; is strikingly like Philoponus'. He contends that a moment may stand at the begin- 

ning of time, with no past time preceding, in exactly the way a point can stand 

at the end of a line segment with nothing beyond. It would clearly be circular 
( reasoning to presuppose that every point must have a line segment on each side 

and then proceed to prove that all lines are infinite. Aristotle's argument from 

the moment is guilty of the same circularity; the argument presupposes its own 
/ conclusion in assuming that every moment must be preceded by time. 270 

Gersonides does not speak of question begging or circular reasoning, but he 

too uses the analogy between space and time. He compares the argument that a 
, moment is always preceded by time and time is therefore eternal, to a possible 

^ argument to the effect that the universe must be infinite since one body comes to 

an end only where it meets another body. Both arguments are a product of the 
( human "imagination," which refuses to allow a point with nothing beyond or a 

moment with nothing before; and the imagination is not a reliable guide in the 
( realm of science. Gersonides appends a lengthy disquisition on the nature of the 

( moment, wherein he shows that not every moment need have the characteristic 

of dividing the past from the future. The upshot is that a moment might serve as 
( the absolute beginning of time just as a point can serve as the absolute beginning 

/ of a line segment. 271 A highly condensed restatement of Gersonides' reply to the 

argument from the nature of the moment is offered by Crcscas, 272 and Abravancl 
i also offers a restatement of Gersonides' reply. 273 

i 

' 2m Epitome of Metaphysics, IV, §4; Tahafut al-Tahafut, I, p. 77. Averrocs' exposition of (he 

^ sophism is extremely subtle. The reason a point can serve as the terminus of a line is, he explains— 

obviously with Aristotle, Physics IV, 13, 222a, 13-14, in mind— that the line is "at rest" and the 
( point at the end of the line enjoys "actual" existence. A moment, by contrast, cannot serve as an 

absolute terminus of time for the reason that time is not at rest, no moment ever exists in actuality, 
( and therefore no moment can ever exist except as subsequent to the past and antecedent to the future. 

Hence, Averrocs concludes, it is a sophism to infer the possibility of time's having its terminus in a 
( moment from the possibility of a line's having its terminus in a point; and Aristotle's argument for 

eternity from the nature of the moment accordingly remains intact. 
1 210 Commentary on Physics, VIII, §983; cf. Summa contra Gentiles, II, chap. 36. Aquinas dis- 

^ misses Averrocs' imputation of sophistical reasoning as follows: The fact that a line is at rest whereas 

! time is ever flowing can have no bearing on the question whether a line or time might have an initial 

| ^ terminus. Therefore, if a line can begin at a point, time should be able to begin at a moment. 

{ 21, Milhamol ha-Shem, VI, i, 21, pp. 388-390. In an indirect reference to Averrocs' strictures 

! ( (above, n. 269), Gersonides, ibid., p. 389, explains that the potentiality associated with the moment 

can have nothing to do with the ability, or inability, of a moment to serve as a beginning of time with 
'.( nothing preceding. 

! r 212 0r ha-Shem, III, i, 3. 

!' ™Mifalot,V,3. 



Proofs of Eternity from Nature of World 



45 



An alternative answer to the argument for eternity from the nature of the 
moment consists in applying the comprehensive response to all arguments from 
the nature of the world. Albcrtus Magnus, 274 Bonaventurc, 275 and Leone Ebreo 276 
grant that the moment always divides past from future in the present state of the 
world; but, they maintain, the moment need not have divided past from future 
during the process of creation. 

(Hi) A further argument for the eternity of time had run: Everything that comes 
into existence does so "in" time; therefore the supposed absolutely first time 
would have to be preceded by another time, namely, the time in which it came 
into existence. 'Abd al-Jabbar, 277 Bonaventure, 278 Gersonides, 279 and Abrava- 
nel, 280 who record the argument, all refute it by rejecting the premise that time 
would have to come into existence in time. 

(e) The response to the argument from the vacuum 

The argument had been that the doctrine of creation ex nihilo implies the prior 
existence of a vacuum in the location the world was to occupy, whereas a vacuum 
is impossible. 281 Aquinas responds that on the assumption of creation ex nihilo, 
"there was no place or space prior to the world," and consequently no vacuum, 
space being created together with the world. 282 Crescas and Abravanel offer the 
same solution. The argument from the vacuum can, they write, be answered by 
understanding that "prior to the world, no dimensions existed, and God created 
them when he created body from nothing." 281 Crcscas makes an additional and 
atypical observation, however. He finds Aristotle's grounds for the impossibility 
of a vacuum unconvincing 284 and is therefore not in the least discomfited by the 
implication of a vacuum prior to creation. Prior to the existence of the world, he 
is quite willing to admit, a vacuum did exist, and in it God created the world. 285 

Gersonides is an adherent of creation who accepts the argument from the 
vacuum. As was seen, he ruled out only those arguments for eternity which rest 
on characteristics of the universe tied to its present state, whereas he endorsed 
arguments which he viewed as resting on universal laws of existence. The argu- 
ment from the vacuum is understood by him to belong to the latter type; for the 



2 'Commentary on II Sentences, d. 1, B, art. 10. 
"'Commentary on II Sentences, d. 1, p. 1, a. 1, q. 2. 
2n Dialoghi d'Amore, p. 239. 
111 K. al-Majmu\ p. 66. 

"^Commentary on II Sentences, d. 1, p. 1, a. 1, q. 2. 
""Milhamot ha-Shem, VI, i, 20, pp. 382-383. 
lm Mifalot, V, 3. 
2 *'Above, p. 27. 

m2 Summa Theologiae, I, 46, I, ad 4. A similar point is made by Saadia, K. al-Amanat, p. 71. 
2 "Or ha-Shem, III, i, 5; Mif'alot, IV, 3. In Abravancl, the sentence ends: "... when he created 
something from nothing." 

284 0r ha-Shem, I, ii, 1. 285 Ibid., Ill, i. 5. 



{ 

' 46 Proofs of Eternity from Nature of World 

I 

^ proposition that objects come into existence in place states a universal charac- 

teristic of physical existence, and the impossibility of a vacuum is also absolute. 

^ Gcrsonidcs accordingly accepts the conclusion that the place now occupied by 

the world must always have contained something, liis position, as has been 

' mentioned, was that matter is eternal and that the world was created out of a 

( preexistent, eternal formless matter. 286 

, (f) The response to the argument from the nature of the celestial bodies 

{ The argument here had been that the process of coming into existence as well 

as the process of destruction consist in a passage from one contrary to another; 
' but whatever substratum the heavens might have cannot be amenable to contrar- 

I ies; consequently, the heavens cannot have come into existence and must be 

eternal. 287 The advocates of creation who undertake to refute the argument employ 
i the standard comprehensive response to arguments from the nature of the world. 

^ They distinguish between natural processes and creation, and maintain: The law 

that things come into existence only from their contraries is a law peculiar to the 



( process of generation within the present state of the world, whereas in creation, 

things may well come into existence in a different manner. This solution is put 
forward in one version or another by Maimonides, 288 Albertus Magnus, 289 Aqui- 

( nas, 290 Gersonides, 291 Aaron ben Elijah, 292 Abravanel, 293 and Leone Ebreo. 294 



i 4. Summary 

I The argument for eternity from the nature of matter rested on the premise that 

everything coming into existence does so from a preexistent substratum or matter; 

( matter too, it followed, could have come into existence only from an already 

^ existing matter; hence an absolute coming into existence of matter, the coming 

into existence of matter from absolutely nothing, is impossible. The proponents 

( of creation reply that the critical premise can be justified solely by induction or 

( 

( 2li6 Milhatnot ha-Shem, VI, i, 17, p. 364. Abravanel, Mif'alot, IV, 3, disputes Gersonides' reasoning. 

( 287 Above, p. 28. 

2m Guide, II, 17. 

I ^ Commentary on II Sentences, d. I, B, art. 10. 

lw Summa Theologiae, I, 46, 1, ad 3. Cf. ibid., ad 2; Summa contra Gentiles, II, 36(1). 
j 29l Milhamot ha-Shem, VI, i, 26. Gersonides adds a statement that, I think, misses the point of 

Aristotle's argument. He contends that not everything is generated from its contrary; forms, for 
( example, are not generated from other forms that are their contraries. Aristotle's contention, however, 

had not been that a given form is generated from the contrary form. His contention had been that a 
' given form is generated in a situation wherein the substratum lacks the given form; and that is the 

^ contrary of the form. See above, p. 28. 

292 '£j Hayyim, chap. 6. 
/ m Mifalot, V, 3. 

2w Dialoghi d'Amore, p. 239. 



/ 



Proofs of Eternity from Nature of World 



47 



analogy, and inductions and analogies are unreliable. Alternatively, the propo- 
nents of creation bring to bear their comprehensive refutation of arguments from 
the nature of the world. They maintain thai although in the present stale of the 
world everything coming into existence does so from a preexistent substratum, 
no preexistent substratum need be assumed for creation. Creation ex nihilo— the 
creation of the world together with its matter — accordingly remains a viable 
hypothesis. 

The argument from the concept of possibility reasoned that everything coming 
into existence is preceded by the possibility of its existence; the possibility of 
existence must inhere in a substratum or matter; matter would therefore have had 
to exist before matter could come into existence; thus the absolute coming into 
existence of matter is impossible. The proponents of creation respond in one of 
three ways. They deny that the possibility of the world's coming into existence 
is to be located in an already existing matter and refer that possibility instead to 
the creator. They construe the possibility of the existence of the world as a 
judgment of the mind and as having no objective existence in the external world. 
Or else they resort to their comprehensive refutation of arguments from the nature 
of the world; and they maintain that in the phenomenon of creation, the possibility 
of existence need not precede actual existence. The upshot is that creation ex 
nihilo is a viable hypothesis. 

The argument from motion showed that the assumption of a beginning of 
motion inescapably involves prior motion; and therefore motion, as well as some 
sort of world undergoing motion, must be eternal. The proponents of creation 
respond by again applying their comprehensive refutation of arguments from the 
nature of the world. They explain that although in the present state of the world, 
motion cannot occur without a prior motion, in the creation of the world as a 
whole, an absolute beginning of motion is conceivable. The world and its motion 
may accordingly have had a beginning. 

The argument from time reasoned that a beginning of time cannot be spoken 
of without implying a prior time, and hence time is eternal. The companion 
argument from the moment reasoned that moments invariably divide past from 
future, and therefore a first moment, with no time and no moments preceding, 
is impossible. Both arguments conclude that since time is eternal, motion and 
some sort of world undergoing motion must also be eternal. In one instance these 
arguments arc countered with the denial that time does involve motion and a 
moving body. If time does not involve motion, the eternity of time may be granted 
without admitting the eternity of the world. The usual response, however, is that 
neither the nature of time nor the nature of the moment implies, in fact, the 
eternity of time. The argument from the nature of time — the contention that there 
is no avoiding the implication of a time prior to the supposed beginning of time — 
is answered by ascribing the implication to deceptive human language; all that 
in truth precedes the beginning of time, the explanation goes, is a supposition of 
time or imaginary time. A supporting consideration is provided by an analogy 



48 



Proofs of Eternity from Nature of World 



between time and space. It was universally acknowledged that the physical world 
is finite and that nothing whatsoever, not even empty space, lies beyond the 
world; by the same token, the adherents of creation submit, time might be finite 
with no time whatsoever preceding the beginning of the world. The argument 
from the nature of the moment is answered by the adherents of creation with the 
aid of another analogy between space and time. They maintain that a moment 
might serve as the absolute beginning of time just as a point can serve as the 
absolute terminus of a line segment. If time need not be eternal and if a first 
moment is possible, then the beginning of motion, and of a world undergoing 
motion, remains a viable hypothesis. 

The argument from the vacuum rested on another principle of Aristotelian 
physics, the impossibility of a vacuum or completely empty space. The argument 
was that the creation of matter is impossible since it would imply the prior 
existence of a vacuum where the world was to come into existence; and a vacuum 
is impossible. The proponents of creation respond that the place the world occu- 
pies was created together with the world, so that no vacuum would have preceded 
creation; and in one instance, the premise that a vacuum is impossible is also 
rejected. It follows once again that creation ex nihilo is a viable hypothesis. 

The argument from the nature of the celestial spheres ran: Everything coming 
into existence does so from its contrary; the circular motion of the celestial spheres 
reveals that the substratum of the spheres is not amenable to contraries; conse- 
quently the spheres and the rest of the physical universe — the existence of which 
is entailed by the existence of the spheres — cannot have come into existence. 
The proponents of creation respond by applying their comprehensive refutation 
of arguments from the nature of the world. The rule that everything coming into 
existence does so from its contrary is, they hold, valid only in the world as it 
exists today and need not be true of the phenomenon of creation. The hypothesis 
of the creation of the world thus remains viable. 



Ill 



Proofs of Eternity from the Nature of 

God 



1 . The proofs 

The proofs for eternity discussed in the previous chapter proceeded from the 
nature of the physical world to the eternity of the physical world, or, in some 
instances, to the eternity merely of the underlying matter of the world. The proofs 
to be discussed in the present chapter take their departure not from the world but 
from its cause; moreover, they cover not merely the physical universe, but the 
nonphysical universe as well. They argue that the cause of the universe being 
such as it is, the entire universe, whether physical or nonphysical, must be eternal. 

Three basic proofs from the nature of the cause of the universe can be differ- 
entiated, and other arguments may be treated as variations of the three. The 
themes of the basic proofs are clearly distinguishable. The first proof argues that 
no given moment, as against any other, could have suggested itself to the creator 
as the proper moment for creating the universe. The second proof argues that the 
cause of the universe must be unchangeable and could not, therefore, have under- 
taken the act of creation after having failed to do so. The third proof argues that 
the cause of the universe possesses certain eternal attributes and that the existence 
of the universe is an expression of those attributes; since the attributes are eternal, 
the universe, which they give rise to, must likewise be eternal. While these three 
lines of reasoning are sufficiently distinct, a difficulty in classification does arise. 
Sometimes a variation on a basic proof or an argument as an individual philos- 
opher happened to formulaic it can plausibly be subsumed under more than one 
basic proof, and the decision to classify it under one rather than another is to an 
extent arbitrary. ' 

The three basic proofs differ in a significant respect that can be brought out 
by considering a possible criticism of any argument for eternity from the nature 
of the cause of the universe. Does not any argument of the sort, it may be asked, 
rest on the unwarranted presupposition that the universe in truth has a cause 

'Cf., for example, below, pp. 59, 64-65. 



49 



50 



Proofs of Eternity from Nature of God 



outside itself? In the case Of the first proof, a satisfactory reply is at hand: The 
standpoint there is, in principle, hypothetical, the intent being to reduce the 
adversary's position to absurdity. A cause of the existence of the universe is not 
presupposed. Rather, the reasoning is that on the assumption of creation, the 
universe would have had to be brought into existence by a cause; 2 the cause 
supposedly bringing the universe into existence would not have been able to 
create it at any given moment; hence the assumption of creation is untenable. 
When stating that the standpoint of the first proof is hypothetical, the qualification 
has to be added that the proof is so in principle, since not every instance of the 
proof keeps the hypothetical standpoint consciously in view. 

In the third proof, the standpoint has plainly shifted away from the hypothetical. 
Presuppositions now are in evidence to the effect that the universe does have a 
cause of its existence, that the cause of the existence of the universe possesses 
certain eternal attributes, and that the attributes express themselves in the exis- 
tence of the universe. The proof is valid only if the presuppositions are granted. 

Most actual instances of the second proof — the proof from the unchangcability 
of the cause of the universe— similarly appear to rest on a presupposition, namely, 
that the universe owes its existence to an unchangeable cause. The second proof 
nevertheless differs from the third in that any argument from the unchangeability 
of the cause of the universe can easily be read or recast in a hypothetical mode. 
On the assumption of creation, the reasoning can be construed, some cause must 
be ultimately responsible for bringing the universe into existence. The cause 
ultimately responsible for bringing the universe into existence would have to be 
immune from change; for were it to undergo change, it would be dependent upon 
whatever produced the change, and it would not, after all, be the ultimately 
responsible cause. 3 And yet, to bring a universe into existence after having failed 
to do so, would constitute change. Creation would thus imply an unchangeable 
cause that undergoes change, and the assumption of creation is self-contradictory. 

The standpoint of the first proof, then, is hypothetical; the second proof can 
easily be read or recast in a hypothetical mode; whereas the standpoint of the 
third is not hypothetical but dogmatic. 

Maimonides, it will be recalled, characterized proofs of eternity from the 
nature of the cause of the world as having been developed from Aristotle's prin- 
ciples by philosophers following Aristotle. 4 The characterization is apt in respect 
to the first and second of the three proofs. A trace of the first proof is to be found 
in Aristotle's physical works, and Aristotle seems to have articulated a version 
of the proof in an early dialogue. 5 The second proof can be seen as a development 
of one of the strands from which Aristotle's argument for the eternity of motion, 

2 For the principle (hat nothing could come into existence without a cause, cf. below, pp. 154 ff. 
3 Scc further, below. 
"Above, p. 10. 
5 Below, p. 52. 



Proofs of Eternity from Nature of God 



51 



discussed in the previous chapter, is woven. 6 Furthermore, the conception of the 
ultimate cause of the universe which is operative in the two proofs accords with 
Aristotle's conception of the prime mover. On the assumption of creation, it is 
argued in both proofs, the ultimate cause of the universe's coming into existence 
would have to be an entity that is unaffected by anything outside itself and that 
is unchanging; and those are traits of Aristotle's prime mover. Maimonides' 
characterization is not, by contrast, accurate as regards the third proof from the 
cause of the universe. The third proof presupposes that there is indeed a being 
from whom the very existence of the universe flows, something quite foreign to 
the spirit of genuine Aristotclianism, where a cause only of the motion of the 
universe is recognized. The third proof would be more accurately characterized 
as Neoplatonic. 

Proclus apparently was the main source or channel through which medieval 
Arabic philosophers received the three proofs for eternity from the cause of the 
world. A list of eighteen proofs of eternity had been drawn up by him, 7 and part, 
if not all, of the list was available in Arabic in the Middle Ages: A medieval 
Arabic translation of the first nine of the eighteen proofs has been discovered, 8 
and in addition ShahrastanT records, in Proclus' name, a paraphrase of seven of 
the nine, together with an eighth proof taken from the second half of Proclus' 
list. 9 The three basic arguments for eternity from the nature of the cause of the 
universe, although not every variation, can be discovered among the proofs of 
Proclus' which have been preserved in Arabic,'" and he may for that reason be 
taken as the mcdicvals' probable source. 

(a ) The argument that nothing could have led a creator to create the universe 
at a particular moment 

The contention that what exists cannot have come into being goes back at least 
to Parmenidcs. In conjunction with other considerations, Parmenides is reported 

''Below, p. 57. 

'Preserved in John Philoponus. De Aclernilale Mundi contra Pmclum. cd. H. Rabc (Leipzig. 
1899). 

"In A. Badawi. ed.. Ncoplalonici apud Arabes (Cairo. 1955), pp. 34-42. Cf. G. Anawati. "Un 
fragmcnl perdu du de aeiernilalc mundi dc Proclus ," Melanges de philosophic grecque offerls a Mgr. 
Dies (Paris, 1950). pp. 21-25. 

9 A'. al-Milal wa-l-Nihal, ed. W. Curcton (London, 1842-1846), pp. 338-340. The eighth proof 
given by Shahrastani is Proclus' thirteenth. In K. Nihaya al-lqdiim, ed. A. Guillaumc (Oxford and 
London, 1934), pp. 45-46, ShahrastanT records, in Proclus' name, three proofs from the tirst half 
of Proclus' list. 

'"I find live of Proclus' proofs— the first, third, fourth, sixteenth, and eighteenth— to be versions 
of the (hrcc proofs discussed in the present chapter. The fifth of Proclus' eighteen proofs is a version 
of the argument from the nature of time; cf. above, p. 24. Most of the remaining are excgctical. 
They undertake to show that Plato's philosophy implies the eternity of the world, and that Plato too, 
despite the apparent sense of the Timaens, believed the world to be eternal in the past as well as the 
future. 



52 



Proofs of Eternity from Nature of God 



to have posed the rhetorical question: "What need could have made that which 
exists, exist later rather than sooner?"" The purport of the question is that since 
nothing could have determined a later, rather than an earlier, moment for the 
existence of the universe, the universe cannot have come into existence at any 
moment whatsoever, but must have existed from eternity. An intimation of the 
same thought can be unearthed in Aristotle. In the course of proving that what 
is indestructible cannot have been generated, Aristotle remarked: "Why . . . after 
not existing for an infinite time, would the thing be generated ... at a particular 
moment?" 12 The argument is known to students of modern philosophy from 
Kant. 13 

Neither the fragment from Parmenides nor the passage in Aristotle makes 
reference to whatever it might be that brings the world into existence. But the 
medieval philosophers as well as Aristotle were certain that nothing could have 
come into existence spontaneously and without a cause. Should the cause respon- 
sible for the supposed creation of the world be taken into account, the argument 
being examined would run: On the assumption of creation, no given moment in 
an undifferentiated eternity could, as distinct from any other moment, have rec- 
ommended itself to the cause bringing the world into existence as the proper time 
for it to create the world; the cause that would, on the assumption of creation, 
have had to create the world could not therefore have acted at any moment 
whatsoever. This train of reasoning, too, was employed by' Aristotle, not however 
in his preserved works but in a lost dialogue. Aristotle, as reported by Cicero, 
argued for the eternity of the world on the grounds that no "new plan" (novo 
consilio) could have arisen which might have occasioned the world's creation. 14 
The mention of a "plan" indicates that the argument has in view not merely the 
moment at which the world would have come into existence on the assumption 
of creation, but also a cause, specifically an intelligent agent, that would have 
had to decide upon and execute the project. Aristotle's meaning must, in other 
words, have been that creation is impossible because there was no moment at 
which the creator might have decided to act. Similar formulations appear else- 
where before the medieval period. Cicero, the source of the passage from the 
forementioned Aristotelian dialogue, represents a member of the Epicurean school 
as asking rhetorically: "Why should the builders of the world suddenly have 
sprung into action after innumerable ages of slumber?" 15 Augustine takes up 
certain unnamed skeptics who had a like predilection for rhetorical questions. 



"H. Dicls, Fragmente der Vorsokraliker (Berlin, 1934-1938), Parmenides, fragment 8. English 
translation in J. Burnet. Early Greek Philosophy (London, 1930). p. 175. 
n De Caelo I. 12, 283a, 1 1-12; cf. Physics VIII, I, 252a, 15-16. 

l1 Kant, Critique of Pure Reason, A427/B455, proof of the antithesis of the first antinomy. The 
proof of the thesis of the first antinomy can be traced to John Philoponus; Of. below, p. 88. 
'"Quoted by Cicero, Academica, II, xxxviii, 1 19. 
"De Nalura Deorum, I, ix, 21. 



Proofs of Eternity from Nature of God 



53 



"Why," the skeptics asked, "did the eternal God choose at a particular point to 
make the heavens and earth, which he had not made previously?" 16 "How did 
the idea of making something come into his mind despite his never having made 
anything before?" 17 

In the Middle Ages, the argument we arc examining — the argument that no 
given moment could, in preference to any other, have served as the proper moment 
for the world's coming into existence — invariably contains a reference to the 
creator. The reasoning invariably is that no given moment could have suggested 
itself to the creator as the proper moment for him to create the world. Avicenna, 
for example, defends the eternity of the world by posing, once again, a rhetorical 
question. He asks: "How within [the stretch of] nonexistence could one time be 
differentiated for [a creator's) not acting and another time for [his) starting |to 
act|? I low might one time differ from another?" IK Instances of the argument that 
no moment could have suggested itself to the creator as the moment for creating 
the world can be found in Ghazali, 19 Maimonides, 20 Albcrtus Magnus, 21 Aqui- 
nas, 22 Crescas, 23 and Joseph Albo. 24 

A distinction can be drawn — and it is admittedly a fine distinction — between 
instances of the argument which do not and a larger number of instances which 
do include an additional clement, the clement of the creator's motive. Avicenna 
and the philosophers just mentioned asked: How could any given moment have 
suggested itself to the creator, in preference to infinite identical moments, as the 
time for creating the world? In the examples to be examined now the question 
is: On the assumption that the creator did choose one moment in preference to 
others, what could possibly have induced him to make the choice? What could 
his motive have been? The theme can already be detected in the passage where 
Aristotle spoke of a "new plan," 25 and in the passage where Augustine's skeptics 
wondered how the "idea" of creating the world could have come into the mind 

"'City of God. XI, 4. 

"Confessions. XI, xxx; cL XI, x. In City of God. XII, 15, Augustine states his own position 
that God was not motivated by "a new plan" (novo . . .consilio) in creating the world. 

"Shifa': Ilahiydt, ed. G. Anawati and S. Zayed (Cairo, 1960), p. 378. 

"Talmfut al-Falasifa, cd. M. Bouyges (Beirut, 1927), I, §28; English translation in Avcrroes 
Taha fit t al-Tahq ful. trans. S. van den Bcrgh (London, 1954), p. 18. 

10 Guide to the Perplexed. II, 14(8). Maimonides includes a point also met elsewhere, namely 
that il is unimaginable that the deity should have remained "idle" for an eternity before creating the 
world. Cf. Cicero, Ue Nalura Deorum. I, ix, 22; Augustine. City of God. II, 18; Simplicius. Com- 
mentary on Physics, ed. H. Diels, Commenlaria in Aristotelem Graeca. Vol. X (Berlin, 1895), p. 
1331. A related notion is that the deity is never "grudging"; cf. below, n. 96. 

"Commentary on II Sentences, in Opera Omnia, cd. A. Borgnct, Vol. XXVII (Paris, 1894), 
d. I, I), art. 10. 

72 Summit contra Gentiles, II, 32(5). 

"Orha-Shem. Ill, i, I. 

24 Iqqarirn, 1, 23. 

"Above, n. 14. 



54 



Proofs of Eternity from Nature of God 



of the creator. 26 An especially clear statement of the theme is provided by Mai- 
monides. He portrays the proponents of eternity as reasoning: "An agent acts at 
one time and not at another because of either preventative factors (mdni') or 
motivating factors (da' in) which occur (tarin) in him. The former bar an agent 
from accomplishing what he wills; the latter lead the agent to will what he 
previously did not will. Since the creator, [who must be absolutely self-sufficient 27 ! 
is subject to neither motivating factors . . . nor preventive factors, ... it is 
impossible for him to act at one time but not at another." 28 

The contention that nothing imaginable could have motivated the creator to 
act at one given moment in preference to others was common. BaqillanI alludes 
to it, and forestalls it, when he stresses that in creating the world "after not having 
done so," the creator was not led to act by a "motivating factor (da' in), ... or 
moving factor, . . . or inducing factor (ba'ith), . . . or disturbing factor, . . . or 
new idea (khatir)? 19 Abu al-Barakat argues that no new "state" (hdl), or "induc- 
ing factor" (ba'ith), or "necessitating factor" (muqtadin) could conceivably have 
led the creator to select "a given time as distinct from what preceded and what 
succeeded, inasmuch as all times were equivalent." The conclusion Abu al-Bar- 
akat draws is that the world must be eternal. 30 With variations in wording, the 
reasoning recurs in Ibn Tufayl, 31 Averrocs, 32 Albertus Magnus, 33 Aquinas, 34 
Fakhr al-DIn al-RazI, 35 Gersonides, 36 Aaron ben Elijah, 37 Crescas, 38 and 
Abravanel. 39 

Proclus' version of the argument we are examining exhibits still a further 
element. His contention is that on the assumption of the creator's acting to create 
the world, the factors inducing him to act at the critical moment would regress 
infinitely. For, if the world were created, the creator would, up to the moment of 
creating the world, have been a "potential creator," and something would have 

26 Above, n. 17. 
27 Cf. below, p. 66. 
2 *Guide. II, 14(6). 

M K. al-Tamhid, cd. R. McCarthy (Beirut, 1957), p. 30. 

i0 K. at-Mu'tabar (Hyderabad, 1939), pp. 33, 43. Abu al-Barakat includes the point that the 
deity would have been "idle" for an eternity. 

}, Hayy b. Yaqdhan, ed. and trans. G. Gauthicr (Beirut, 1936), Arabic text, p. 82; French 
translation, pp. 62-63; English translation with pagination of the Arabic indicated: Hayy Ibn Yaqztm, 
trans. L. Goodman (New York, 1972). 

i2 K. al-Kashf; ed. M. Mueller (Munich, 1859), p. 31; German translation with pagination of 
the Arabic indicated: Philosophic und Theologie von Averroes, trans. M. Mueller (Munich, 1875). 

"Physics, in Opera Omnia, ed. A. Borgnet, Vol. Ill (Paris, 1890), VIII, tr. I , chap. 1 1(6). 

'"DePolentia, q. 3, art. 17(13). 

"Muhassal (Cairo, 1905), p. 91. 

M Milhamol ha-Shem (Leipzig, 1866), VI, i, 18, p. 371. 

37 'Es Hayyim, ed. F. Delitzsch (Leipzig, 1841), chap. 7. 

3 "Or ha-Shem, III, i, 4. 

}9 Mif'alot (Venice, 1562),VI, 1(1), quoting what appears to be Ghazali, Tahafut al-Ealasifa, 
1, §5. 



Proofs of Eternity from Nature of God 



55 



had to "activate" him. But the activating factor would, before inducing the creator 
to create the world, have been a "potential" activating factor, and hence would 
have stood in need of a prior factor to activate it as well. And the prior factor 
would in its turn also have had to be activated. The supposition of creation 
therefore leads to an infinite regress of factors activating the creator. Since an 
infinite regress of causes is absurd, 40 the world, Proclus concludes, must be 
eternal. 41 

Argumentation of the same type is alluded to and forestalled, as in a previous 
instance, 42 by BaqillanI. He explains that the creator cannot be understood to 
have created the world "for a reason" (li-'illa), since the "reason" for the world's 
creation would have been absent up until the moment of creation and would only 
then have come into existence. Its coming into existence at the moment of creation 
could only have been due to a different reason; and the reasons for the creation 
of the world would regress infinitely. 43 'Abd al-Jabbar records an argument for 
eternity according to which the creator's becoming active after having been inac- 
tive would have to be due to a factor (ma'nd). The factor's being present and 
operative precisely at the moment of creation would have to be due to another 
factor, and it, in turn, to yet another. Therefore, the argument concludes, the 
assumption of creation implies an infinite regress of causes, and is absurd. 44 
Averrocs defends the doctrine of eternity, writing: If the creator be assumed to 
have "acted at a given time and not at another, some cause would have had to 
assign him the one state as distinct from the other." But then an additional cause 
would have to be responsible for the cause's being present at the critical moment, 
and so forth ad infinitum. The thesis of creation is consequently absurd. 45 In a 
separate passage, Averroes contends in a similar vein that to assume a beginning 
of motion in the universe would imply an infinite regress of movers, which is 
absurd. 46 That argument is later recorded by Gersonides. 47 

The Kalam writers were fond of the notion that when events might take more 
than a single course, something must tip the scales between the equivalent pos- 
sibilities and determine the course that is taken. 48 The notion of lipping the scales 
is employed now by several Kalam writers in restating their opponents' argument 
that creation would imply an infinite regress of motivating factors. Ghazali has 

""See below, pp. 241, 337. 

"'Quoted by Philoponus, De Aelernilate, pp. 42-43. 

42 Above, n. 29. ' 

"K. al-Tamhid, pp. 31-32. 

M Slmrh al-Usfd (Cairo. 1965). p. 115. 

45 AT. al-Kashf, p. 30. 

^Epitome of Metaphysics , cd. and trans. C. Quiros Rodriguez (Madrid, 1919), IV, §3; German 
translation: Die Epitome der Metaphysik des Averroes, trans. S. van den Bcrgh (Leiden. 1924), pp. 
105-106. The argument is closely related to the argument from motion, discussed in the previous 
chapter. 

"Milhamot ha-Shem, VI. i, 3, p. 299. 
4 *CI'. below, p. 162. 



56 



Proofs of Eternity from Nature of God 



the adherents of eternity reason: The world's being produced at a given moment 
would require the presence at the given moment of a "factor tipping the scales" 
(murajjih) in favor of the creator's acting. The presence of the factor at just the 
required moment would demand another factor to tip the scales in favor of it, 
and so on ad infinitum. But an infinite regress of causes is inadmissible; hence 
the assumption of creation is inadmissible. 49 Similar formulations are recorded 
by Fakhr al-DIn al-RazI, 50 Amid!, 51 and Ijl. 52 And RazI and Iji describe the 
argument that creation would imply an infinite regress of factors tipping the scales 
as the "pillar" of the adherents of eternity. 

Resume 

A progression can be discerned. As far back as Parmenides we find the thought 
that no given moment in the undifferentiated stretch of eternity could, in pref- 
erence to any other, have lent itself to the world's coming into existence. The 
argument thereupon developed that no given moment could have recommended 
itself to the creator as the proper moment for him to create the world. A number 
of medieval philosophers offer that version, usually through the medium of a 
rhetorical question. An even larger number offer a version into which the element 
of motivation is introduced. The thinking here is that no imaginable motive 
could have induced the creator to act and create the world at one given moment 
rather than another. Finally, the element of an infinite regress is added by Proclus 
and by a line of medieval philosophers who are probably dependent directly or 
indirectly on him. They contend that the factors, or reasons, or causes, or factors 
tipping the scales, which would have led the creator to create the world at a given 
moment would regress infinitely. Since an infinite regress of causes is impossible, 
the creation of the world is impossible as well. 

(b) The argument from the unchangeability of the cause of the universe 

Should an unchangeable cause of the existence of the universe be presup- 
posed, an argument for eternity can be framed which runs: An unchangeable 
cause is known to be responsible for the existence of the universe. But an 
unchangeable cause would not pass from a state of inaction to a state of action. 
The cause of the universe cannot, therefore, have acted to bring its effect into 
existence after having failed to do so, and the universe must be eternal. The 
argument can also be put in a hypothetical form, although, it must be confessed, 
actual instances of the hypothetical form are not the rule. Put hypothetically, the 
argument would go: On the assumption of creation, the universe must have been 
brought into existence by a cause. The ultimate and true cause of the coming into 



49 Tahaful al-Falasifa, I, §§3, 5; English translation, p. 1. 
50 K. al-Arba'in (Hyderabad, 1934), p. 42. 
5 'Ghdya al-Maram (Cairo, 1971), p. 265. 
52 Mawaqif (Cairo, 1907), VII, pp. 228-229. 



Proofs of Eternity from Nature of God 



57 



existence of the universe would have to be unchangeable. For if the immediate 
cause of the universe's coming into existence underwent change, it would be 
dependent on whatever produced the change; if what produced the change under- 
went change, it would in its turn be dependent upon something else; and unless 
an unchangeable cause should be reached, nothing would be truly and ultimately 
responsible for the universe's coming into existence. 5 '' The assumption of creation 
implies, then, an ultimate cause that is unchangeable. Yet the assumption equally 
implies that the cause ultimately responsible for the universe's coming into exis- 
tence did change inasmuch as it passed from a state of not acting to a state of 
acting. Creation thus embodies a self-contradiction and is untenable, and the 
universe cannot have been created but must be eternal. 

The argument from the unchangeability of the cause of the universe, especially 
when formulated hypothetically, resembles and could well have been suggested 
by a strand in Aristotle's proof of the eternity of motion, a proof that will be 
recalled from the preceding chapter. At one stage of that proof Aristotle reduced 
the assumption of an absolute beginning of motion to absurdity by looking at the 
relationship between the cause producing, and the body performing, the supposed 
first motion or change. The relationship between what produces and what per- 
forms the supposed first motion would, he reasoned, have had to change before 
the motion could occur; and the motion or change assumed to be absolutely first 
would, consequently, not be first after all. 54 The argument we arc presently 
examining, the argument from the unchangeability of the cause of the universe, 
looks, for its part, not at the relationship between the cause of the universe and 
the universe, but at the cause alone. The argument reduces the assumption of a 
beginning of the universe to absurdity by showing that the assumption would 
imply a change in the cause ultimately responsible for creation, whereas anything 
subject to change could not after all be an ultimately responsible cause. 

The argument from the unchangeability of the cause of the universe can also 
be seen as an extension of the proof examined under the previous heading, the 
argument that no given moment could have suggested itself to the creator as the 
proper moment to act. In the argument from the unchangeability of the cause of 
the universe, the impossibility of an infinite regress can explain — although other 
explanations arc possible too 55 — why the cause ultimately responsible for the 
existence of the universe would have to be unchangeable: The unchangeability 
of the ultimate cause must be posited in order to avoid an infinite regress of 
changeable causes. Creation is thereupon found to be untenable because it would 
imply that a cause which must be unchangeable nevertheless changes. In the final 
version of the proof examined under the previous heading, the contention was, 
more simply, that a cause of the universe could not act to bring the world into 

"Cf. below, p. 337. 
54 Above, pp. 19-20. 

"Sec, for example, Proclus' argument, immediately below. 



58 



Proofs of Eternity from Nature of God 



existence at a given moment because the factors inducing the cause to act would 
regress infinitely. 56 The present argument goes beyond the previous one in only 
a single detail. The unchangeability of the cause of the universe is explicit here 
but not there. 

The argument for eternity from the unchangeability of the cause of the universe 
is found in Proclus, who defended the eternity of the universe by reasoning: The 
cause of the universe must be "immovable." For were it subject to motion, it 
would pass from a state wherein it was "imperfect" to a state wherein it was 
"perfect," which is inconceivable in regard to the highest being. And it would 
moreover "stand in need of time," which is inconceivable in the being that is the 
cause of everything outside itself, including time. But if "an agent is immovable, 
it is unchangeable; and if unchangeable, it . . . [cannot| pass . . . from [a state 
ofj not acting to [a state of] acting. . . . Therefore, if something ... is an 
immovable cause of something else . . . it is so eternally . . . ; and if the cause 
of the universe is immovable ... the universe must be eternal." 57 In the Middle 
Ages, the argument for eternity from the unchangeability of the cause of the 
universe was employed or recorded by Aviccnna, 58 Ibn Hazm, 59 ShahrastanI in 
his paraphrase of Proclus, 60 ShahrastanI in his account of Aristotle's philosophy, 61 
and Averroes. Averroes' formulation runs: The assumption of creation entails 
either that things come into existence spontaneously, which is absurd; or else that 
the agent bringing the world into existence underwent "change and hence stood 
in need of an agent apart from itself to bring about the change." On the latter 
alternative, the second agent would stand in need of a further agent; and we 
would be left in the end with no "first agent" who is responsible for the world's 
coming into existence. The creation of the world is accordingly untenable and 
the world must be eternal. 62 

Sometimes the concepts of potentiality and actuality are called into play. On 
the hypothesis of creation, the reasoning goes, the creator would have passed 
from the state of being a potential creator to the state of being an actual creator; 
such a transition would be a change; but change is impossible for the cause of 



56 Above, pp. 54-55. 

"Quoted by Philoponus, De Aelernitale, pp. 55-56. This is only part of Proclus' argument, 
and the rest is examined under the next heading. The notion that the unchangeability of the deity 
implies the eternity of the world also appears in other proofs of Proclus'; see Philoponus, ibid., 
pp. 42, 604. 

,s Shifa': llahiyat, p. 376. 

,9 A\ al-Faslfi al-Milal, (Cairo, 1964), I, p. 20; Spanish translation: Abenhazam de Cdrdoha 
y su Historia Critica de las Ideas Religiosas, trans. M. Asm Palacios, Vol. II (Madrid, 1928), 
pp. 113 ff. 

m K. al-Milal wa-l-Nihal, p. 339; K. Nihaya al-Iqdam, p. 46. 
61 K. al-Milal wa-l-Nihal, p. 320. 
62 Tahafut al-Tahafut, 1, p. 8. 



Proofs of Eternity from Nature of God 



59 



the universe; and hence creation is impossible. That formulation is recorded by 
ShahrastanI, 64 Albcrtus Magnus, 65 Bonavcnture, 66 Aquinas, 67 Gcrsonidcs, 68 Aaron 
ben Elijah, 69 Crcscas, 70 and Abravancl. 71 Maimonides, 72 and Aquinas 73 record 
a truncated version, which runs: If the world had been created, the creator would 
have passed from a state of potentiality to a state of actuality; something would 
have had to bring about the transition; and nothing could possible bring about a 
transition from potentiality to actuality in the cause of the universe. The reason 
why no transition from potentiality to actuality can be brought about in the cause 
of the universe is not stated. Very likely, the intended reason is that the cause of 
the universe must be unchangeable. 74 But the reason could equally be — as in the 
final version of the proof examined under the previous heading 75 — that the acti- 
vating factors would regress infinitely. And, since the eternal actuality of the 
ultimate cause had been established by Aristotle and other philosophers, 76 the 
reason could also be that the cause of the universe is always in a state of pure 
actuality. In fact, these ostensibly different reasons arc interconnected. The 
impossibility of an infinite regress of activating factors can serve as grounds both 
for the unchangeability, and the eternal actuality of the ultimate cause; 77 and the 
unchangeability of the ultimate cause both entails, and is entailed by, its state of 
eternal actuality. 78 All three principles — the impossibility of an infinite regress, 
the unchangeability of the cause of the universe, and the eternal state of actuality 
of the cause of the universe — were commonplace in the Middle Ages, and any 
or all of them could easily have been taken for granted. 



6, See the argument of Proclus' discussed under the previous heading, above, p. 55. Philoponus, 
De Aeterniiaie, p. 82, quotes an argument from another work of Proclus' to the effect that if the 
world were created, the creator would have been in a state of potentiality, and hence imperfect, before 
becoming actual. 

64 K. al-Milal wa-l-Nihal, p. 339; K. Nihaya al-lqdum, p. 45. ShahrastanI is paraphrasing the 
argument of Proclus' which I included under the previous heading, above, pp. 54-55. 
"Physics. VIII, ir. I. chap. II. 
"'Commentary on II Sentences, d. 1, p. I, a. 1, q. 2. 
" De Potentia. q. 3, art. 17(12). 
'"Millumwt ha-Shem, VI, i, 24, p. 393. 
<f> 'Es Hayyim, chap. 7. 
™Or ha-Shcm. Ill, i, I. 
n Mifaht. VI, 1. 
12 Guide, II, 14(5). 
ls Summa contra Gentiles, II, 32(2). 
74 Aquinas gives this reason elsewhere; above, n. 67. 
"Above, p. 55. 

76 Aristotle. Metaphysics XII, 6, 1071b. 20; below, p. 347. 

77 To avoid an infinite regress of causes, a first cause must be posited which is neither changeable 
nor subject to a transition from potentiality to actuality. 

7 *A cause that is not subject to change will always be in its slate of actuality; and anything that 
is eternally actual will never change. 



60 



Proofs of Eternity from Nature of God 



The argumentation thus far has been that the ultimate cause of the universe 
must, in general, be unchangeable and consequently could not pass from inaction 
to action. Several variations are in evidence as well, each of which directs its 
attention to a respect wherein the cause of the universe would have to be 
unchangeable. One variation appears in Augustine, 79 Proclus, 80 Maimonides, 81 
Aaron ben Elijah, 82 and Gersonides. 83 Proclus argues, and the others record the 
argument, that the decision to create the world at a particular moment would 
constitute a change in God's will. Inasmuch, however, as God's will must be 
identical with his essence, 84 and his essence is unchangeable, his will must be 
unchangeable. God consequently could not have decided to create the world after 
having failed to make the decision, and the world must be eternal. 

According to another variation, reported by 'Abd al-Jabbar, creation would 
imply a change in God's "knowledge," hence "a change in his state"; for the 
creator would have spent an eternity without knowing that the world exists, 
whereupon, at the moment of creation, he would acquire a new item of knowl- 
edge, the knowledge that the world does exist. But since God is not changeable, 
his knowledge is not changeable. Therefore the external object of his knowledge, 
the universe, is unchangeable and must have existed forever. 85 

Still another variation was probably inspired by Aristotle's point that the 
assumption of a beginning of motion would imply a prior change in the relation- 
ship between the cause producing, and the object performing, the supposed abso- 
lutely first motion. 86 Crescas records an argument concerned not with the change 
of relationship which would precede creation but with the change of relationship 
which would result. The argument is that on the assumption of creation, the cause 
of the universe, or the deity, would enter a new relationship; for before creation 
he would not have had a relation to the universe, whereas subsequently he would. 
But a change in the creator's relationship to the world would entail a change in 
himself, which is an impossibility. Consequently, creation is an impossibility. 87 
The argument is repeated by Abravanel. 88 



79 CityofGod, XII, 18. 

80 Quoted by Philoponus, De Aelernilate, p. 560. 

81 Guide. II, 14(6); 18. 

82, £j Hayyim, chap. 7. 

"Mithamot ha-Shem, VI, i, 18, p. 377. 

"The reason is that the essence of the first cause must be absolutely simple. Cf. Plotinus, 
Enneads, VI, 8, 13. On the question whether an Aristotelian or Plotinian deity can properly be 
described as having will, sec E. Zeller, Die Philosophie der Griechen, Vol. II, Part 2 (4th cd.; 
Leipzig, 1921), pp. 368-370; III. 2 (5th ed.; Leipzig, 1923), pp. 539-540. 

s> Sharhal-Usul, p. 117. 'Abd al-Jabbar docs not explain how the authors of the argument would 
harmonize the unchangeability of God's knowledge with the constant changes in the objects of his 
knowledge within an eternal universe. Reasoning similar to that recorded by 'Abd al-Jabbar appears 
in Khayyat, K. al-Intisar, ed. and trans. A. Nader (Beirut, 1957), §71. 

86 Above, p. 19. 

"Or ha-Shem, III, i, 1. '"Mifalot, VI, 1. 



Proofs of Eternity from Nature of God 



61 



In sum, it was argued that creation would imply a change in the cause of the 
universe, or a change in his will, or in his knowledge, or in his relationship to 
the world; but a change of any sort is impossible in the cause of the universe; 
therefore the creation of the world is impossible. 

Arguments from God's eternal attributes* 9 

An argument for eternity from the attribute of divine goodness 91 ' was known 
to Augustine. He portrays the Stoics as defending their peculiar theory of an 
eternity of world cycles — as distinct from the eternity of a single world — on the 
grounds that God's "goodness" could never have been "inoperative"; 91 since 
God's goodness must always have been in operation, the reasoning went, there 
must always have existed an expression of his goodness, that is to say, a world. 
A similar argument appears in a more fully developed form in Proclus, who 
characterizes it as the "most convincing . . . demonstration" 92 of the eternity of 
the world. The original Greek text of the passage in Proclus is lost, but Philo- 
ponus' refutation has been preserved in the original Greek, and a medieval Arabic 
translation of Proclus has also been preserved. Philoponus' refutation indicates 
that the argument had originally taken God's goodness— together with his power — 
as its premise. 93 In the medieval Arabic translation, however, a small change has 
been made and the proof rests on God's beneficence {jud), together with His 
power. 

Proclus, as refracted through the Arabic translation, lays down the proposition 
that "when an agent docs not act, he fails to act either because he does not wish 
to or because he is unable to." In the issue at hand, the world's coming into 
existence, it is impossible that the agent should not have wished to act. For it is 
unimaginable that the supreme being should be merely "sometimes beneficent 
(jawad), sometimes not." He undoubtedly is "eternally beneficent"; and, being 
eternally beneficent, he must eternally "wish the universe to resemble him," to 
exist and be good, even as he exists and is good. 94 The impossibility that the 
deity should not have wished to bring the world into existence is matched by the 

89 Thc sense in which the deity can have attributes, and the proper manner of construing terms 
predicated of him, were of course perennial and much debated questions. Cf. H. Wolfson, Philo, 
Vol. II (Cambridge, Mass., 1948), pp. 149-164; idem, Studies in the History of Philosophy and 
Religion, Vol. I (Cambridge, Mass., 1973), pp. 98-169; idem, Philosophy of the Kalam (Cambridge, 
Mass., 1976), pp. 112-234. 

'"'That God's goodness is the source of the existence of things outside of God was a tenet of 
Platonic and Ncoplatonic philosophy. Sec Plato, Republic, 509B; Plotinus, Enneads, V, 5, 10; 13; 
Arabic paraphrase: Plotinus apud Arahes , cd. A. Badawi (Cairo, 1955), p. 182; Proclus, Elements 
of Theology, cd. E. Dodds (Oxford, 1963), §12. Aristotle, for his part, characterized the prime mover 
as supremely good, Metaphysics XII, 7 and 9. 

""City of God, XII, 18. 

92 In Badawi, Neoplatonici apud Arabes, p. 34. 
91 De Aelernilate, p. 13, et passim. 
94 Cf. Plato, Timaeus, 29E. 



62 



Proofs of Eternity from Nature of God 



impossibility that he should be incapable of accomplishing what he wishes. Inas- 
much as the deity must have wished eternally to bring the universe into existence 
and must, moreover, be eternally able to effect what he wishes, he must have 
brought the universe into existence eternally.""' Unlike the Stoic arguments in 
Augustine's report, Proclus' argument concludes that God's beneficence — which 
perforce produces something as similar to God as possible — must give rise to a 
single eternal universe, as distinct from an eternal succession of universes. Vir- 
tually all medieval arguments from the eternal divine attributes likewise conclu- 
ded that God's attributes must give rise to a single eternal universe. 

In Arabic, the argument for eternity from the beneficence of the first cause is 
common. The adherents of eternity, 'Abd Jabbar explains, "base their thesis" 
on, among other things, the premise "that the creator is eternally beneficent." If 
God were not eternally beneficent, "he would be sometimes beneficent and at 
other times not beneficent . . . [but) niggardly, 96 [which is unimaginablc|. Once 
the foregoing is established, the eternal existence of the world follows." 97 The 
argument from God's "beneficence" is alluded to by Avicenna, 9 ^ and is found in 
Ibn Hazm, 99 in Shahrastanl's paraphrase of Proclus, 100 in Fakhr al-DIn al-RazT, 101 
AmidI, 102 Ijl , 1 03 and Leone Ebreo. 104 Aquinas records an argument in the same 
vein which infers the eternity of the world from God's "most perfect" and "infinite 
. . . goodness." 105 

The proof for eternity from the deity's eternal attributes, like the other proofs 
for eternity, exhibits a number of variations. In each variation an eternal attribute 
of God is shown to express itself eternally in the existence of the world, God's 
knowledge, his wisdom, his perfection in general, and his character as Lord of 
the universe, all being brought into play as grounds for eternity. 'Abd al-Jabbar, 
for example, takes up an argument to the effect that God cannot have brought 
the world into existence through the "inducement of need" (da'i al-haja), since 
he is self-sufficient and immune from need. 106 God could only have brought the 
universe into existence through the "inducement of wisdom" (da'i al-hikma). 
God's wisdom comprises "his knowledge of the goodness of the world and his 
knowledge of the benefit others will derive from it. That [knowledge] is firm 



"in Badawi, Neoplalonici apud Arabes, p. 34. 

96 For the doctrine that the deity is free of "grudging" (<p06v<><;), see Plato, Phaedrus, 247A; 
Timaeus, 29E; Aristotle, Metaphysics I, 2, 983a, 2; Philoponus, De Aelernitale, p. 13. 
97 K. al-Majmu' ft al-muhit bi-l Takltf, ed. J. Houben (Beirut, 1965), p. 66. 
m Shifa: Ilahiyat, p. 380. 

99 K. al-Fasl fi al-Milal, I, p. 20. The passage is quoted below, p. 64. 
lm K. al-Milal wa-l-Nihal, p. 339; K. Nihaya al-Iqdam, p. 45. 
m K. al-Arba'm, p. 50. 
' 02 Ghaya al-Maram, p. 266. 
""Mawaqif. VII, p. 230. 
""Dialoghid'Amore, p. 238. 

""Summa contra Gentiles, II, 32(7). Cf. De Potentia. q. 3, art. 17 (1, 14). 

106 This part of the argument is related to the arguments discussed above, pp. 53-54. 



Proofs of Eternity from Nature of God 



63 



through eternity. And therefore the existence of the world is necessary through 
eternity." 107 Here the contention is that God's eternal wisdom and knowledge 
would require him to produce the world eternally. Aquinas has a kindred argu- 
ment, which intertwines and perhaps confuses two notions, the notion that God's 
eternal knowledge is the cause of the universe and the notion that true knowledge 
mirrors objects in the external world. Aquinas' version reads: "God is the cause 
of things through his knowledge." Knowledge is "relative" to the thing known; 1 " 8 
for knowledge and the object of knowledge "exist together by nature," true knowl- 
edge occurring solely when something actually exists to serve as its object. God's 
knowledge of what is brought into existence through his knowledge accordingly 
requires, for it to be true knowledge, that the things in question actually exist. 
And inasmuch as "God's knowledge is eternal, things apparently are produced 
by him from eternity." 109 

The notion that true knowledge must mirror objects in the external world also 
plays a role in a passage where Gcrsonides, followed by Crescas and Abravancl, 
speculates about the underlying considerations leading Aristotle to the doctrine 
of eternity. Gcrsonides observes that "an intelligible thought corresponding to 
no object outside the mind would seem necessarily to be false." But "God is the 
[ intelligible | order \nimus\ of the universe"; that is to say, God's thought, which 
is identical with his essence, comprises a mental representation of the universe. 
Aristotle concluded herefrom, in Gcrsonides' reconstruction, that the actual and 
external order of the universe must exist whenever God and his thought — which 
comprises the intelligible order of the universe — exist. God consequently "could 
not exist without the universe." and since God is eternal, the universe must be 
eternal. 110 

Another ground adduced for the eternity of the world was the broad attribute 
of divine perfection. Avcrrocs writes that since God is perfect, he cannot "fail 
to perform the superior act and perform instead the lesser act; for that would be 
a defect" in him. It would surely be the "greatest defect" in "the eternal agent, 
whose existence and act are [in fact] unlimited, . . . were his act to be limited 
and finite." God's act, the causation he exercises vis-a-vis the universe, 111 must, 
therefore, be infinite in duration, and the universe, which is the product of his 
causation, must be eternal." 2 

Eternity could, as has been seen, be inferred from the attribute of divine 
perfection or from the attribute of divine wisdom. Maimonidcs, followed by 



""Slwrhal-Ustil. p. 116. 

" m Cf. Aristotle, Topics IV, 1, 121a, 1. 

,m De Potentia. q. 3, art. 17(19). 

""Gcrsonides, Milhamot ha-Shem. VI, i, 3, 302; Crescas, Or ha-Shem, III, i. 1 (very much 
abbreviated); Abravancl. Mif'aht, VI, 1 (copying from Crescas). 

"'In Avcrrocs' view, God's causation is primarily in the realm of motion; but, as a cause ol 
motion. God is also the cause of the existence of the universe. Sec below, pp. 325-326, 341. 

" 2 Tahafi<t al-Tahafut, I, p. 96. 



64 



Proofs of Eternity from Nature of God 



Albcrtus Magnus, shows how eternity might be inferred from the two attributes 
conjointly. The adherents of eternity, Maimonides reports, argued that since 
"God's acts are perfect and contain no defect," and God makes "everything as 
perfect as can be," our "universe must be the most perfect possible." The existence 
of the world flows, moreover, from God's wisdom, which undoubtedly maxi- 
mizes the perfection of the world. God's wisdom is, in its turn, "identical with 
his essence" and hence "eternal," so that the maximum measure of existence it 
could bring forth is eternal existence. The world, as the most perfect possible 
product of God's eternal wisdom, must likewise be eternal." 3 

Proclus' argument for eternity from the deity's goodness or beneficence had 
cited the attribute of divine power side by side with the attribute of divine good- 
ness. Proclus had reasoned that since God is good and beneficent, he could not 
fail to wish the existence of the world; and since God is powerful he could not 
fail to accomplish what he wishes." 4 lbn Hazm knows of a singularly spare 
argument combining not two but three divine attributes: "[The proponents of 
eternity] affirm that the cause of the creator's act is his beneficence, wisdom, and 
power; he is always beneficent, wise, and powerful; and seeing that the cause of 
the world always exists, the world must always exist."" 5 Coincidentally, Aquinas 
too records an argument combining the attributes of divine knowledge, divine 
power, and divine goodness. It runs: God is "not ignorant" — not lacking in 
knowledge — and hence he must know how to produce the world from eternity. 
He is "not impotent" — not lacking in power — and hence he must be capable of 
producing the world from eternity. He is not "envious" — not lacking in 
goodness — and hence he must want to produce the world from eternity. The world 
must consequently be eternal." 6 

One more variation of the proof from the eternity of God's attributes was 
suggested to Aquinas by a passage in which Augustine described God as "Lord 
from eternity." 117 It might be argued, Aquinas observes, that since God is eter- 
nally the "Lord," he must eternally have subjects with respect to whom he can 
be designated as the Lord; and a universe must therefore always exist." 8 

The foregoing are arguments for eternity from God's eternal attributes. A 
further recurring argument consists in the inference of the eternity of the universe 
from the deity's being a cause, but the appropriate location of that argument in 
the classification scheme I am observing is not clear-cut. The argument posits 
that the deity is, by his nature, eternally a cause by virtue of itself or that he is 

" 3 Maimonides, Guide, II, 14(7); Albcrtus Magnus, Physics, VIII, tr. I, chap. 11. 

""Above, pp. 61-62. 

'"A:. al-Faslftal-Milat, I, 20. 

" 6 De Potenlia, q. 3, art. 17(22). 

" 7 Augustine, City of God, XII, 16; the quotation is not exact. The editions of De Potenlia refer 
to Augustine, De Trinitale, V, 16, but I could not find an appropriate passage there. 
us De Potenlia, q. 3, art. 17(21). 



Proofs of Eternity from Nature of God 



65 



an eternally actual cause. Both those terms arc convertible into unchangeable 
cause, 1 19 and the argument from the deity's being a cause may accordingly be as- 
simil.atcd to the argument from the unchangcability of the cause of the universe. 
The deity's being a cause by his very nature may, however, also be regarded as 
a divine attribute, and the argument may accordingly be treated as an added 
variation of the proof from the eternal attributes of God. It can thus be plausibly 
assigned to either of two headings. In any event — and wherever the argument is 
best classified — the earliest instances I could discover are two passages in Proclus 
where the deity's character as an eternal cause appears not independently, but 
interwoven with other considerations. 

The first passage happens to be the one in which Proclus proved that the cause 
of the universe must be unchangeable. 120 In the course of working out his proof 
for eternity from the unchangcability of the cause of the universe, Proclus explains 
that what is unchangeable cannot be a cause "sometimes." If it is a cause at all, 
it must be "a cause eternally"; and the cause of the universe, which is indeed 
known to be unchangeable, must eternally possess the character of being a cause. 
Having come this far, Proclus proceeds to the inference that is of interest here. 
He lays down the rule that an eternal cause is perforce "a cause of something 
eternal." Inasmuch as the cause of the universe is an eternal cause, and an eternal 
cause is a cause of something eternal, the cause of the universe must, Proclus 
concludes, brings forth an eternal universe. 121 

The second pertinent passage in Proclus is the one in which he has been seen 
to contend that creation would imply an infinite regress of factors activating the 
creator. 122 The complete argument there takes the form of a dilemma. The creator, 
Proclus writes, was either once potential, whereupon he became actual, or else 
he was eternally actual. If the former alternative were correct and the creator 
had once been potential, the factors required to activate him would, as was seen 
earlier, regress infinitely, which is impossible. The latter alternative, then, remains, 
and the cause of the universe must be eternally actual. But Aristotle showed that 
"when . . . the cause is actual, the effect is likewise actual." 123 Given Aristotle's 
proposition together with the proposition that "the creator is an eternally actual 
creator," the conclusion "ensues that his effect is likewise eternally actual." An 
eternally actual cause of the universe must bring an actual universe into existence 
eternally. 124 

In the Middle Ages, lbn Hazm recorded an argument resembling the second 
of the two arguments just quoted from Proclus. lbn Hazm's version takes the 

""if something should exist exclusively by virtue of itself, there would be no factor that could 
bring it into a new condition; and similarly, if something is purely actual and completely free of 
potentiality, it could never be brought into a new condition. 

l20 Abovc, p. 58. 

l2 'Quotcd by Philoponus, De Aeternitale, p. 56. l22 Cf. above, pp. 54-55. 

'"Physics II, 3, 195b, 17-18; 28. Also cf. Plotinus, Enneads, IV, 5, 7. 
l24 Quoted by Philoponus, De Aeternitale, pp. 42-43. 



66 



Proofs of Eternity from Nature of God 



form, again, of a dilemma: On the assumption of creation, the creator brought 
the universe into existence either "by reason of himself (li-annahu) or . . . by 
reason of a cause |distinct from himself]." The latter alternative is inadmissible 
since it would imply an infinite regress of causes inducing the creator to act. 12 '' 
The former alternative therefore remains, and the creator must have acted to 
produce the world "by reason of himself." But a "cause is inseparable from its 
effect"; that is to say, when the cause is present, its effect is present. Inasmuch 
as the creator is a cause by reason of himself, he is a cause as long as he exists, 
and inasmuch as he exists eternally, he is a cause eternally. Consequently, his 
effect, the universe, must also exist eternally. 126 Fakhr al-DIn al-RazT and Ijl 
record the same argument, formulated with the aid of the same dilemma. 127 

Avicenna is more direct. The most expeditious way of settling the issue of 
creation and eternity is, he maintains, through understanding that the deity is a 
cause by virtue of itself. "A cause by virtue of itself produces its effect by 
necessity {awjaba)? so that whenever it exists, it acts. "If such a cause exists 
eternally, it acts to produce its effect eternally." Therefore, the eternal cause of 
the universe must produce an eternal universe. 128 Avcrroes drops the qualification 
"by virtue of itself," and argues: When the agent "is eternal," in other words, 
when something is eternally an agent, "its action must be . . . eternal, and its 
effects must be eternal." 129 

Bonaventure and Aquinas know of an argument based on the deity's being a 
"sufficient" cause, a term akin to cause by reason of itself or cause by virtue of 
itself. A sufficient cause is such that it does not require the aid of auxiliary causes 
or conditions to accomplish its ends, which means that as soon as a "sufficient 
cause is given, its effect is given." Since God is a sufficient cause and since he 
exists eternally, his effect, the conclusion goes, must also exist eternally. 13 " Besides 
the argument from God's being a sufficient cause, Aquinas records the contention 
that God's "action . . . must be eternal" because it "is identical with his sub- 
stance, which is eternal." The conclusion again is that "the effect" of God's 
action, the universe, must be eternal. 131 

Fakhr al-Din al-RazI has yet another version of the argument from the deity's 
being an eternal cause. He describes the proponents of eternity as maintaining, 
in astonishingly good Kalam style, that to be an agent (mu'aththir) is not a 
"negative attribute" but an "eternal . . . positive attribute added to the essence." 
To be an agent, the thinking continues, is a "relative attribute," that is, an attribute 
implying a correlative. 132 Since being an agent is both a positive and a relative 



l25 Cf. above, p. 55. 

126 K. al-FaslJTal-Milal. I, p. 9. 

,21 K. al-Arha'in, pp. 41-42; Mawaqif, VII, 228-229. 

™Shifa: Ilahiyat, p. 373. U "K. al-Kashf, p. 30. 

""Bonaventure, Commentary on II Sentences, in Opera Omnia, Vol. II (Quaracchi, 1882), d. 1, 
p. 1 , a. 1 , q. 2; Aquinas, Summa Theologiae, I, 46, art. 1 , obj. 9; Summa contra Gentiles, II, 32(3). 
"'Summa Theologiae, I, 46, art. 1 , obj. 10; De Polentia, q. 3, art. 17(6). 
" 2 Cf. above, p. 63. 



Proofs of Eternity from Nature of God 



67 



attribute, its presence implies the existence of a positive correlative, that is, an 
actually existing corresponding effect. God's being an agent eternally thus implies 
the eternal existence of his effect, the universe. 133 

Gcrsonides, followed by Crcscas and Abravanel, provides still another version. 
Medieval Aristotelians had been divided on the question whether the first cause 
of the universe is, or is not, identical with the incorporeal mover of the outermost 
celestial sphere. Some philosophers, most notably Avicenna, understood that the 
first cause of the universe — the deity — is a being beyond the incorporeal mover 
of the outermost sphere; 134 but others understood that the highest being in exis- 
tence, the deity, is the mover of the outermost sphere, and that beyond the movers 
of the spheres nothing further exists. The second position was espoused by Avcr- 
roes, 135 and Gcrsonides naturally enough takes it to be the genuine position of 
Aristotle. Gcrsonides now speculates that one of the underlying considerations 
which led Aristotle to his belief in eternity must have been his position regarding 
the mover of the outer sphere. Aristotle's reasoning, as reconstructed by Gcr- 
sonides, was that since the deity is by his nature the cause of the motion of the 
outermost celestial sphere, he could not exist without a sphere to move. The 
eternal existence of the first cause would imply the eternal existence of the sphere 
moved by him, and hence the eternity of the rest of the world as well, seeing 
that the existence of the rest of the world is entailed by the existence of the 
sphere. 136 

2. Replies to proofs from the nature of the cause of the universe 

The medieval advocates of creation have no single comprehensive response to 
arguments for eternity from the nature of the cause of the universe similar to 
their comprehensive response to arguments from the nature of the world itself. 137 
What we do find is that each of the three proofs from the nature of the cause of 
the universe elicits its own set of responses, and that some ancillary motifs recur 
in answers to more than one proof. It is, I think, not unfair to add that the 
responses to arguments from the nature of the cause arc less satisfactory than 
were the responses to arguments from the nature of the world. 13 " 

"'A'. al-Arha'in, p. 49. 

" 4 Cf. Avicenna, Shi/a': llahiyat, pp. 392-393. 401. 

"'""Cr Avcrroes. Long Onmnentary on Metaphysics, XII, comm. 44. 

'"■Gcrsonides. Milhamot ha-Shem, VI, i, 3. p. 302; Crcscas. O- ha-Shein, III, i. I; Abravanel. 
Mif'alot, VI, 1. Crcscas and Abravanel give the argument in a very abbreviated form. 
" 7 Cf. above, pp. 30-33. 

""This is recognized by Crcscas. Or ha-Shem. Ill, i, 4 and 5, and Abravanel. Mif'alot. VI, 3 
(end). As a means of escaping the arguments for eternity from the nature of the cause, Crcscas 
inclines towards, and Abravanel embraces, an old rabbinic theory (Genesis Rahhah, ix. 2) according 
to which God continually creates and destroys a succession of worlds. Sec Or ha-Shem, III. i. 5; 
Mif'alot. VII, 3 and 5 (Abravanel distinguishes his own theory of successive worlds from Crcscas' 
theory). It is difficult to sec, however, just how the theory of successive worlds answers the lirst two 
of the three proofs. Leone librco, Diuloghi d'Amore, p. 252, attributes a theory of successive worlds 
lo Plato. 



68 



Proofs of Eternity from Nature of God 



(a) Responses to the argument that nothing could have led a creator 
to create the world at a particular moment 

The proof, as will be recalled, had several variations or stages. The proponents 
of eternity argued that no given moment in an undifferentiated eternity could, in 
preference to any other, have suggested itself as the proper moment for the creator 
to create the universe; that there could be no imaginable motive for the creator 
to create the universe at a given moment; that the factors motivating the creator 
to create the universe at a given moment would regress infinitely, whereas an 
infinite regress is impossible. Medieval responses to the proof in its several 
variations rest largely on the thesis that the creator brought the world into exis- 
tence through an exercise of will, the central thesis then being buttressed with 
secondary considerations. The central thesis as well as the secondary consider- 
ations can be discovered in Augustine and John Philoponus, either in contexts 
where the two were replying to the present proof, or in contexts where they were 
replying to other proofs for eternity. 

Augustine takes up the question how the creator might have created the world 
at a given moment in time as distinct from the infinite other identical moments 
when the world could have been created. He does not so much answer the question 
as dismiss it, branding it as a sophism on the grounds that there was "no time" 
before creation, and consequently the act of creation cannot legitimately be spo- 
ken of as having occurred in time at all. 139 As for the question what the new 
factor might have been which induced the creator to act after not having acted 
for an eternity, Augustine denies that any factor need be posited. No new cir- 
cumstances, he submits, induced the creator to create the world at the moment 
when he did. For God did not create the world "in accordance with a new | plan | 
. . . but rather in accordance with an eternal plan," 140 and in accordance with 
"one and the same eternal and immutable will." 141 From all eternity God's will 
determined the moment at which the world should come into existence. Through 
one and the same eternal plan, and one and the same eternal act of will, creatures 
"previously were not, as long as they were not"; "and they thereupon were," 142 
at what Augustine — despite his insistence that creation did not strictly occur in 
time at all — allows himself to style a given "time." 143 In addition to thus respond- 
ing to the arguments on their own merits, Augustine voices an ad hominem ani- 
madversion. It consists in applying the analogy between space and time, an 
analogy that advocates of creation were earlier seen to utilize when responding 
to arguments for eternity from the nature of the world. 144 Augustine is addressing 



' "Ci/v of God, XI, 6; Confessions, XI, 30. 
""'CilvofGod, XII, 15. 
""Ibid., 18. 
,42 Ibid. 

,4, Ibid., XI, 5. 
l44 Abovc, pp. 41-42. 



Proofs of Eternity from Nature of God 



69 



proponents of eternity who recognize a cause of the existence of the universe, 
and he contends that they, at least, do not have the right to pose difficulties 
regarding the moment chosen for creation. Such proponents of eternity must 
acknowledge that the deity has determined a location in space for the existence 
of the finite physical world in preference to the infinite alternative places where 
the world might exist. Consistency would therefore require that they acknowledge 
the deity's ability to determine a moment in eternity for the beginning of the 
existence of the world in preference to the infinite other moments when the world 
might have begun to exist. 145 

Augustine's position, in sum, is that the world cannot legitimately be described 
as having been created at a moment in time. No new factor motivated the creator 
to act, for the previous nonexistence, and subsequent existence of the world were 
determined eternally and immutably through an eternal divine plan and an eternal 
act of the divine will. Proponents of eternity who recognize a cause of the exis- 
tence of the universe cannot pose difficulties about the deity's ability to create 
the world at one given moment in preference to other possible moments, since 
they must admit something analogous, namely the deity's ability to produce the 
world — albeit eternally — at a given place in preference to the other possible 
places where the world might exist. 

John Philoponus was confident that the creation of the world was not merely 
defensible, but demonstrable, 146 and his response to the proof for eternity we arc 
examining builds on his demonstration of creation. It must be recognized, Phil- 
oponus stresses, that creation can be demonstrated and that the universe must 
have conic into existence at some moment. It must further be recognized that 
questions such as "why the universe did not come into existence earlier" can be 
asked about every moment at which the world might have been created. Since 
the universe is known to have come into existence at one moment or another, 
and since questions can be raised about whatever is assumed to have been the 
moment of creation, the questions, Philoponus holds, arc pointless and can be 
dismissed. 147 

Like Augustine, Philoponus explains that creation took place in conformity 
with an eternal decision of God's eternal will. Philoponus is answering Proclus, 
who, for his part, had also affirmed the eternity of God's will, but had inferred 
therefrom, in another proof for eternity than the one we are now examining, that 
the world is eternal. Proclus' reasoning had been that since the world exists by 
virtue of God's will, which is eternal, it must likewise be eternal. 148 In response, 
Philoponus insists upon the distinction between God's eternally willing that a 
thing should exist and his willing that the thing should exist eternally. Although 



"City of God, XI, 5. 
'"Sec below, p. 93. 
l7 De Aetcrriiralc. pp. 11-12. 
' K Cf. above, p. 60. 



70 



Proofs of Eternity from Nature of God 



the adherents of creation do maintain that God willed the existence of the world 
from eternity, this is not identical with, nor docs it entail, God's having willed 
the eternal existence of the world. 149 To support the proposition that an eternal 
will can determine nonctcmal events, Philoponus develops an ad Imminent device, 
different from Augustine's, lie is addressing philosophers who acknowledge, as 
Philoponus understands Proclus to have done, that the divine will is responsible 
for whatever occurs in the universe. Those philosophers concede, in effect, that 
through a "single . . .simple. . . act of willing" 150 the deity wills the occurrence 
of myriad nonctcrnal and changing events at given determinate moments in the 
course of history. How, Philoponus marvels, can the same philosophers shrink 
from the proposition that God eternally willed the occurrence of an additional 
noneternal event at a given moment, to wit, the coming into existence of the 
world as a whole'? 15 ' 

The following motifs, to recapitulate, are brought into play by Augustine and 
Philoponus either in answer specifically to the proof for eternity we are presently 
examining or in answer to other proofs. No moment in time can strictly be 
described as having been chosen for creation, inasmuch as there was no time 
before creation. The creation of the world at a given moment came about in 
accordance with an eternal decision on the part of the creator. To will something 
eternally is not equivalent to willing its eternal existence. Once creation is con- 
clusively demonstrated, the question why one moment rather than another was 
chosen becomes pointless. At any rate, proponents of eternity who recognize a 
cause of the existence of the universe perforce acknowledge that the deity is 
responsible for the existence of the world in a given location, as distinct from 
every other possible location where the world might exist. They can therefore 
hardly balk at the thesis that the creator is capable of choosing a given moment 
for the existence of the universe. Proponents of eternity who, moreover, recognize 
a being possessed of will which is the cause of everything occurring in the 
universe perforce acknowledge that an eternal unchanging will can decide upon 
the occurrence of given events at given moments; they therefore cannot balk at 
the thesis that an eternal unchanging will could decide upon the creation of the 
world as well at a given moment. In the Middle Ages, these motifs recur in 
varying combinations. 

The question "Why did God not create the world prior to the time" when he 
did is taken up by Saadia. And Saadia responds: Prior to creation "there was no 
time that could be asked about; furthermore, it is of the character of a voluntary 
agent to act when he wishes." 152 

,w De Aelernilaic, p. 566. 
"°Ibid., p. 568; cf. p. 81. 
"'Ibid., pp. 566-567; 580. 

'"Saadia, K. al-Amanal wa-l-l'liqadat, cd. S. Landaucr (Leiden, 1880), p. 73; English trans- 
lation with pagination of Arabic indicated: Book of Beliefs and Opinions, trans. S. Rosenblatt (New 
Haven, 1948). 



Proofs of Eternity from Nature of God 



71 



Ghazali addresses the rhetorical question what could differentiate one specific 
time from earlier and later times as the moment for creation, and responds by 
referring to the creator's will. It is, he too explains, of the nature of will, and not 
merely divine will, precisely to differentiate between things that are similar in 
every respect. Hence by the exercise of sheer will, God could select one from 
among an infinite number of identical moments and designate it as the moment 
for creation. To support his assertion, Ghazali has recourse to the analogy between 
time and space. He observes, much as Augustine had done, that adherents of 
eternity who recognize a cause of the existence of the universe cannot avoid 
acknowledging wholly arbitrary determinations in the spatial realm. The location 
of the north and south poles 153 at a given pair of points on the celestial sphere 
is, in the example he offers, wholly arbitrary, seeing that any other pair of opposite 
points would be as suitable. Since Ghazali's adversaries recognize that the first 
cause, the deity, did arbitrarily determine a pair of points for the poles, how, he 
remonstrates, can they question the deity's competence arbitrarily to determine 
a moment for the world to begin to exist'? 154 

Ghazali also responds to the argument that the factor "tipping the scales" in 
favor of creation at a particular moment would have to have the scales tipped in 
its favor by a previous factor, and the latter by yet another factor, ad infinitum. 
An eternal will, he maintains, is capable of determining that something should 
remain nonexistent up to a certain stage and only thereupon begin to exist; con- 
sequently, no factor need have tipped the scales at the critical moment. 155 To 
support the assertion, Ghazali observes, as Philoponus had done, that his adver- 
saries also trace myriad temporal events in the universe back to the eternal first 
cause of the universe. To grant that the eternal first cause is responsible for the 
occurrence of temporal events, even through intermediate causes, is to grant that 
the first cause is ultimately responsible for the moment at which each event 
occurs. How then, Ghazali again remonstrates, can the proponents of eternity 
question the ability of the eternal first cause to bring about one additional non- 
ctcrnal event at a determinate moment, namely the coming into existence of the 
entire universe'?' 56 

Ghazali's contentions reappear in Abu al-Barakat, who cites them in the name 
of "the creationists." The creationists, Abu al-Barakat informs us, answer their 
adversaries by maintaining that the divine will 15 ' is fully competent to differ- 
entiate between things similar in every respect, such as identical moments; that 
the adherent of eternity too has to acknowledge arbitrary determinations made 
by God's will in the realm of space and should not shrink from arbitrary deter- 
minations by God's will in the realm of time; that an eternal will can decide upon 

"■'That is to say. the points around which the celestial spheres rotate. 
'™T<ihaful al-l-ulusifu. I. §§30, 38; English translation, pp. 14, 24. 
'"Ibid., §8; English translation, p. 3. 
" 6 lbid., §47; English translation, p. 32. 
'"Ghazali stated this of will in general. 



72 



Proofs of Eternity from Nature of God 



a noneternal effect; that the adherent of eternity traces events occurring at given 
moments back to the divine will and consequently should not question the cotr- 
petencc of the divine will to bring the entire universe into existence at a given 
moment.' 38 Abu al-Barakat was not in the least swayed by any of these consid- 
erations and his belief in the eternity of the world remained unshaken. 159 

Kalam writers coming after Ghazali were, however, confident that the proof 
for eternity under discussion had been refuted. Fakhr al-DIn al-RazI takes up the 
question: What tipped the scales in favor of creation at a particular moment? He 
responds that no factor had to "tip the scales," for the selection of the moment 
was accomplished exclusively by God's "will." The arbitrary selection of a moment 
for creation, RazI adds, is analogous to the arbitrary selection of a location for 
the stars on the celestial spheres. Inasmuch as adherents of eternity acknowledge 
the ability of God's will, in the spatial realm, to assign given locations to the 



\ stars in preference to other possible locations, they cannot deny the ability of his 

will, in the temporal realm, to fix upon a given moment for creation in preference 
to others. 160 The creation of the world at a given moment is, RazI observes in 
( another work, analogous to the emergence of sundry individual events at count- 

' less moments through history. The adherents of eternity recognize that sundry 

temporal events do flow from the unchanging state of actuality of the first cause, 
(' albeit through intermediate causes; and they recognize that no factor standing 

behind the first cause tips the scales to activate him, since to assume such a factor 
' would lead to an infinite regress. The adherents of eternity thereby acknowledge, 

( in effect, that the first cause is ultimately responsible for the moment at which 

each temporal event occurs. How then can they question the ability of the first 
cause to determine a moment for one additional event, the coming into existence 
( of the entire universe? 161 

The argument that creation would involve an infinite regress of factors "tipping 
^ the scales" is also dealt with by AmidI and Iji. AmidI counters by reference to 

( God's eternal "will," which determined "that the nonexistence of the world would 

extend up to" a particular moment and that the world would "come into existence 
' at the moment when it did come into existence." 162 Iji writes: "A voluntary agent" 

( is capable of "tipping the scales ... in favor of one of two [equipollent] alter- 

natives within his power by pure will, with no need of" an "inducing factor . . . 
' added to himself." And at any rate, the adherent of creation is in no worse a 

( predicament than adherents of eternity who admit that the universe has a cause. 



The latter acknowledge that individual "things daily coming into existence" are 
' traceable back to the eternal cause of the universe and that no factor intervenes 

( 



( ""AT. al-Mu'tabar, III, pp. 33, 43-44. 

'"Below, p. 75. 
/ l60 Muhassal, p. 91. 

"■'K. al-Arba'in, p. 51. 
( '"Ghaya al-Maram, p. 268. 



( 



Proofs of Eternity from Nature of God 



73 



to tip the scales and induce the eternal cause to act. They can hardly scruple at 
attributing the coming into existence of one more thing — the world in its entirety — 
to an eternal cause, without the intervention of a factor tipping the scales."''' 

Aquinas similarly espouses the theory, which he could have learned from cither 
Augustine or Ghazali, that an eternal will is capable of deciding from eternity 
when a noneternal product should come into existence. No new factor, he writes, 
need have activated the creator at the moment of creation. For God produced the 
world through his "thought and will," which are eternal, and they both eternally 
determined that the world should come into existence at the moment when it 
did. 164 

Maimonidcs affirms that the creator could act at a given moment without any 
"inducement" (da" in), prompting him to act, nor any "preventive factor" (muni'), 
blocking his action up to the moment he acted; and Maimonidcs defends that 
position by setting forth the difference between two kinds of voluntary agent. 
The voluntary agents that act or fail to act because of inducing or preventative 
factors arc, he explains, agents that seek to attain "a purpose . . . external to the 
will itself." For example, a man is induced to build a house by factors outside 
his will, such as the weather or environment; and he may be prevented from 
building the house by factors outside himself, such as the absence of building 
materials. By contrast, the agent who acts "exclusively by will" is not affected 
by external factors. Nothing external to the will motivates such an agent, and 
nothing outside him prevents him from acting. He acts or refrains from acting 
with absolute autonomy and exclusively as his will dictates; consequently, without 
external inducement, and exclusively as his will determined, he would be com- 
petent to decide upon the moment for creation. 165 Aaron ben Elijah 166 responds 
to the argument that nothing could have induced the creator to create the world 
at a given moment in the same manner as Maimonidcs. 

In the foregoing instances, medieval adherents of creation explain away diffi- 
culties regarding the moment chosen for creation. Several medieval adherents of 
creation take the more radical step of branding their adversaries' questions as 
illegitimate and disallowing the questions outright. 

Part of Augustine's response to difficulties regarding the moment chosen for 
creation was the consideration that before creation time did not exist. 167 The 
thought reappears when TusI — who hardly could have had any link, direct or 
indirect, to Augustine — addresses the familiar question: How could God have 
settled upon a moment for creation in preference to the infinite other possible 
moments when creation might have occurred? To handle the matter properly, TusI 

'"'Mawaqif, VII, 229-230. 
* M Summa contra Gentiles, II, 35. 
' 65 Guide, II, 18(2). 
,66 £'s Hayyim, chap. 7. 
""Above, p. 68. 



74 



Proofs of Eternity from Nature of God 



asserts, one must understand that time did not precede the existence of the world. 
The supposed time before creation was "imaginary," not real, and any "differ- 
entiation" between moments before creation is equally "imaginary." In no true 
sense was a choice made between moments, and it is illegitimate even to ask 
how God could have settled upon a particular moment. " ,9 

Aquinas, in one passage, combines the consideration that time did not exist 
before creation with an application of the analogy between time and space. The 
question why the creator should have chosen a given moment for creation, rather 
than any other, is compared by him to the question why God should have chosen 
to locate the world at a given spot in space, rather than any other. Outside flic 
world, Aquinas notes, there is no space, and to speak of places outside the world 
where the world might have been located is illegitimate. Similarly, there was no 
time before creation, and to speak of moments before the existence of the world 
when the world might have come into existence is no less illegitimate. Questions 
regarding the time before creation, like those regarding space outside the world, 
may be dismissed. 170 

In Gersonides, the thesis that God's eternal will eternally determined the moment 
for creation is joined with a consideration found earlier in Philoponus, 171 but 
presumably reached by Gersonides independently. Gersonides treats the familiar 
problem: What might have led God to create the world at the moment when he 
did? And the familiar solution is offered: The choice of a moment may be ascribed 
to God's eternal unchanging will. But Gersonides understood creation to be 
demonstrable and not merely defensible, which means that if a world was to 
exist, it could not be eternal but had to be created. 172 That being so, he continues, 
the question why creation occurred at a given moment rather than any other loses 
legitimacy. Inasmuch as the world must have come into existence at some moment, 
and inasmuch as the question regarding the moment of creation could have been 
raised regarding whatever moment God should have chosen, the question may 
be dismissed as pointless. 173 

Thus far the adherents of creation have cither answered or disallowed their 
adversaries' questions regarding the possibility of a moment's being selected for 
creation. Kalam writers perceived that the thrusts and parries over the possibility 
of a moment's being selected for creation could also be turned to a constructive 
end. The defense of the doctrine of creation led almost invariably to the creator's 
will; an exercise of sheer will had to be assumed in order to explain the selection 



,68 Cf. above, pp. 41-42. 

""Gloss to Muhassal, p. 92. A similar position is taken by Leibniz in A collection of Papers 
which passed between the Late Mr. Leibnitz and Dr. Clarke (London, 1717), III, §6; V, §55-60. 
""Summa contra Gentiles, II, 35(5). 
'"Above, p. 69. 
l72 Cf. below, pp. 209-211. 

m Milhamotha-Shem, VI, i, 18, p. 377; cf. ibid., 24. pp. 395-397. Albcrlus Magnus answers 
questions about the factors inducing the creator to act (cf. above, n. 33) in a similar fashion; his 
response is that an eternal world is impossible. See Albcrtus, Physics. VIII, tr. I, chap. 14. 



Proofs of Eternity from Nature of God 



75 



of a moment for creation. As long as advocates of creation adopted a purely 
defensive attitude, will in the creator remained nothing more than an assumption 
offered in support of another assumption, the assumption of creation. The Kalam 
school was, however, certain that creation can be demonstrated conclusively. 
And once creation is demonstrated, will in the creator may be seen as its dem- 
onstrated corollary. For if the world is known to have been created, one out of 
an infinite number of identical moments is known to have been selected for 
creation, and the creator, who selected the moment, is known to be possessed of 
will. The realization that will in the creator is implied by the doctrine of creation 
was incorporated into the standard Kalam procedure lor proving the existence of 
God, which consisted in demonstrating the creation of the world and inferring 
the existence of God from creation. 174 Inasmuch as the selection of a moment 
for creation would imply will in the creator, Kalam thinkers could, alter having 
proved creation, infer not merely the existence of a creator but the existence of 
a creator possessed of will. The procedure of proving creation and inferring 
therefrom not merely a deity, but a deity possessed of will, is employed by 
BaqillanI, 175 'Abd al-Jabbar, 176 Juwaynl, 177 Ghazali, 178 RazT, 179 Amid!, 180 and 
Ijl."" 

The theory that a moment for creation could have been selected by God s 
eternal will elicited surrejoinders from Abu al-Barakat and Avcrroes. Abu al- 
Barakat makes two points, only the second of which he regards as decisive. 
Adherents of eternity can, he writes, surrejoin by noting that whenever an agent 
resolves to do something in the future, "another resolution Cuzinut) or exercise 
of will" is required to activate the agent when the awaited moment arrives. The 
theory of eternal will does not, therefore, exempt the proponents of creation from 
explaining what might have induced the creator to create the world when the 
moment did arrive for him to act. The second point made by Abu al-Barakat — 
and, in his view, the decisive one — is that effects are never deferred fortuitously. 
Whether a cause acts "through will or without will," its effect is delayed only 
because of "a deficiency in the causality," that is to say, cither because "knowl- 
edge . . . strength, . . . will," happen to be absent, or because "preventative 
factors" happen to interfere. Inasmuch as the deity lacks nothing needed for action 
and nothing ever hinders him from acting, his action can nowise be delayed. 
Whatever he wills must be accomplished forthwith, and if he should will the 
existence of a universe, he could not help but bring a universe into existence 
immediately. 1X2 



l74 Cf. above, p. 2; below, p. 154. 
n -K. al-Tamhid. p. 27. 
m Sharh al-Usftl. p. 120. 
I77 K. al-lrshdd. pp. 28-29. 

""K. al-lqtisad)t al-rtiqad (Ankara. 1962), pp. 101-102. 
m K. al-Arahii'in. pp. 147-148, with nuances. 

m 'Ghaya al-Maram, p. 45. ""Mawaqif. VIII, pp. 82-83 

'" 2 K. al-Mu'talxir. Ill, pp. 34-35; cf. ibid., pp. 47-48. 



76 



Proofs of Eternity from Nature of God 



Both these points reappear in Averroes, who reasons: On the assumption that 
an eternal will predetermined the moment for the world to come into existence, 
a new "resolution ('aim), . . . which did not exist prior to the moment in ques- 
tion," would still be required in order that the eternal agent and his eternal will 
should be activated. The theory of eternal will consequently docs not exempt the 
adherents of creation from the necessity of explaining what induced the creator 
to act at the moment when he did. 183 And in another work Averroes writes: No 
"delay" is conceivable between a "voluntary agent's resolution to act" and his 
acting, 184 nor is any delay conceivable between the "agent's acting" and the 
appearance of the effect. Immediately upon God's resolution to produce the world, 
the world must have existed; and if God's resolution to produce the world is 
eternal, the world cannot have come into existence later but must have existed 
from all eternity. 185 

(b) Responses to the argument that creation would imply change in the creator 
The second proof for eternity from the nature of the cause of the universe turns 
on two propositions: the proposition that a cause ultimately responsible for the 
existence of the universe would have to be unchangeable, and the proposition 
that should creation be assumed, the cause ultimately responsible for the existence 
of the universe would have undergone change in passing from inaction to action. 
If, on the one hand, an ultimate cause of the universe must be unchangeable, 
whereas creation would, on the other hand, involve change in the ultimate cause, 
the assumption of creation is untenable.""' The advocates of creation usually 
respond by inverting the reasoning. They agree that the ultimate cause, or deity, 
must be unchangeable. But the line they take is that since the deity must be 
unchangeable, the act of creating the world would not, in his case, have consti- 
tuted a change. 

That response is offered by Ibn Hazm, who imparts to it a Kalam tinge. 
"Alteration," Ibn Hazm writes, "consists in something's coming into existence 
in the subject of alteration which was not there before, with the result that the 
subject exchanges its [previous] attribute ... for another." A situation of the sort 
is not possible in God inasmuch as "he is never the subject of an attribute." God 
accordingly acts or does not act "by virtue of his essence" and without change. 187 
Bonaventure employs similar language, although historical links with Ibn Hazm 
cannot easily be supposed. When the deity produces something new, Bonaventure 



m K. al-Kashf, p. 31. 

184 Averroes obviously is using the expression resolution lo act in a sense that excludes a voluntary 
agent's resolving to do what he is not capable of. 

'"Tahafut al-Tahdfut, 1, pp. 7-8; the English translation is not satisfactory. What appears here 
in Averroes as a surrejoinder is recorded by Aquinas as an independent argument for eternity; Summit 
contra Gentiles, II, 32(4). 

'""Above, pp. 56-57. 

187 K. al-Fasl ft al-Milal, I, p. 20. 



Proofs of Eternity from Nature of God 



77 



maintains, nothing "is added" to him which was not there before; consequently, 
the act of creation would not constitute change in God. 188 

The same approach is also taken by Maimonides except that, as would be 
expected, Aristotelian terminology is now used. Maimonides deals with the ver- 
sion of the argument where creation is shown to involve a transition from poten- 
tiality to actuality in the agent responsible for creation, while the cause ultimately 
responsible for creation could not undergo such a transition. 189 And Maimonides 
replies: A transition from potentiality to actuality occurs, as all Aristotelians 
recognize, 1911 solely in material objects. Therefore when an incorporeal agent 
does produce something after not having done so, no passage from potentiality 
to actuality occurs in him. 191 Maimonides' solution is repeated by Aaron ben 
Elijah. 192 Another formulation is provided by Aquinas, but the burden is still the 
same. "A new divine effect docs not," in Aquinas' words, "signify a new action 
in God; for God's action is identical with his essence" and hence is, like his 
essence, eternally unchangeable. God could therefore produce a new effect with- 
out undergoing change. 193 

One variation of the proof for eternity from the unchangcability of the cause 
of the universe concerned itself with will, the contention being that a change 
specifically in the creator's will would be implied by creation. 194 The advocates 
of creation respond here as they do to the basic argument. Maimonides and, 
following him, Aaron ben Elijah lay down the proposition that voluntary agents 
undergo a change of will only when they are corporeal; for only corporeal vol- 
untary agents act "to attain an external end," hence their wills alone arc moved 
by external inducing and preventative factors. 195 By contrast, the will of an 
incorporeal agent, which seeks no external end, is immune from change. So far, 
of course, the advocates of eternity would concur; and they would conclude that 
an agent who is unaffected by circumstances outside himself and who is immune 
from change could not possibly decide at a particular moment to bring the world 
into existence. Maimonides and Aaron, however, conclude not that creation is 
impossible, but rather that creation would involve no change in the will of an 
incorporeal agent: Inasmuch as the will of an incorporeal agent is not moved by 
external factors, when — or if — an incorporeal agent did begin to bring the uni- 
verse into existence, his action would be accompanied by no change of will. 196 



188 'Commentary on II Sentences, d. 1, p. 1, a. 1, q. 2, ad 6. 

Above, pp. 58-59. 
'""See Zcllcr, Die Philosophic dcr Griechen, II, 2, pp. 318-323. 
'"'Guide, II, 18(1). 
m 'Es Hayyim, chap. 7 

m Summa contra Gentiles, II, 35(1). Cf. De Potentia, q. 3, art. 17 (ad 12). 
,94 Above, p. 60. 
l95 Cf. above, p. 73. 

l96 Maimonidcs, Guide, II, 18(2); Aaron ben Elijah, 'Es Hayyim, chap. 7. 



78 



Proofs of Eternity from Nature of God 



Another version of the argument from the unchangcability of the deity reasoned 
that creation would involve a ehangc of relationship in the creator. 1 '" Crescas, 
who gave that version, responds: "A new relationship does not constitute a change 
of essence [in the case] of an incorporeal being." 19 * An incorporeal, as distinct 
from a corporeal, agent could as a consequence bring something new into exis- 
tence and thereby enter a new relationship without undergoing any change in 
himself. 

The line taken by these adherents of creation betrays more than a whiff of 
question begging. Their adversaries had pressed the plausible thought that if an 
agent never undergoes change he cannot conceivably do something he did not 
do before. In responding, the adherents of creation insist, as firmly as their 
opponents, on the deity's immunity from change. And they nowise reveal how 
an agent might begin doing something he previously did not do without changing. 
They circumvent the straightforward and obvious conclusion that the deity never 
passes from the state of not creating the world to the state of creating it, and they 
assert instead that when the deity docs pass from the one state to the other, he 
does so without changing. 

Philoponus and, in the Middle Ages, ShahrastanI and Gersonides make a dif- 
ferent response, which — whether cogent or not — docs come to grips with their 
opponents' argument. All three belong to the circle of philosophers who arc 
confident that the creation of the world can be demonstrated definitively. Taking 
the truth of creation and the impossibility of eternity as a premise, they explain: 
The deity eternally wills the creation of the world (Philoponus and Gersonides) 
and is eternally an "actual creator" (ShahrastanI), but since an eternal world is 
intrinsically impossible, the world cannot possibly be produced by God from 
eternity. The world comes into existence as soon as it can, and the soonest it can 
come into existence is in the finite past. God, thus, eternally and unchangeably 
wills the existence of the universe and is eternally in the state of creating it; yet 
all that his eternal willing of the universe and his eternal act of creation gives 
rise to is a noneternal universe. 199 

The response to one final version of the argument from the unchangcability of 
the deity may be mentioned. 'Abd al-Jabbar recorded a version according to 
which creation is inadmissible because it would imply a change in the creator's 
knowledge; the creator would previously not know, and subsequently would 
know, that the universe is in existence. 200 The response 'Abd al-Jabbar makes is 
that knowledge of something not yet existent is nowise different from knowledge 



'"Above, p. 60. 

198 Or ha-Shem, III, i, 2. The printed editions are corrupt and have to be corrected with the aid 
of manuscripts of the text. 

'"Philoponus, De Aelernitale, p. 81. taken together with p. 8; ShahrastanI. A'. Nihaya al-lqdam, 
pp. 47-48; Gersonides, Milhamot ha-Shem, VI, i, 18, p. 377. 

200 Above, p. 60. 



Proofs of Eternity from Nature of God 



79 



of the same thing when it exists. The assumption of creation docs not, therefore, 
imply that any change occurred in the creator's knowledge. 20 ' 

(c) Responses to the arguments for eternity from God's eternal attributes 

The contention here had been that God's attributes — including his character 
of being a cause by his very nature — express themselves in the existence of the 
universe, and therefore the eternity of his attributes entails the eternity of the 
universe. 202 Rebuttals take three forms. The adherents of creation deny that the 
attributes concerned are eternal attributes of God; they maintain that although 
God may possess the attributes eternally, the attributes do not necessarily express 
themselves by producing a universe eternally; or else they undertake to show that 
it would be intrinsically impossible for God's attributes to express themselves in 
an eternal universe. 

The first form of response is to be found in 'Abd al-Jabbar, who refuses to 
"concede that God is eternally beneficent"; and if God is not eternally beneficent, 
the eternity of the world plainly cannot be inferred from his beneficence. 201 

Shahrastani's thinking is more subtle. He denies that beneficence is an "essen- 
tial attribute added to the essence," and interprets it instead "as an attribute of 
action," or "a relation." The statement that God is beneficent means no more than 
that God is an "agent"; and the statement that God "is sometimes beneficent, 
sometimes not," means no more than that "he is sometimes an agent, sometimes 
not an agent." Since the term agent and the attendant term beneficent have their 
reference not in God but in what he has produced, in the relation of the universe 
to God, nothing in God's essence could disclose whether he is eternally, or merely 
sometimes, beneficent. The question whether God is, or is not, eternally benef- 
icent and eternally an agent cannot, then, serve as a starting point for settling 
the issue of eternity and creation. On the contrary, that question is itself the 
"locus of dispute" between the proponents of creation and the proponents of 
eternity, and can be resolved only through settling the issue of eternity and 
creation on its own merits. The eternity of God's beneficence hinges on the 
eternity of the universe and not vice versa. 204 

In a similar vein, Amid! refuses to interpret divine beneficence as an "attribute 
of perfection" in God, that is to say, as an attribute pertaining to the perfection 
of God's essence and consequently coeternal therewith. Beneficence is "an attrib- 
ute of action," reducible to God's "being an agent" with no "end" or "profit" in 
view. Amid! gives two reasons for not construing divine beneficence as an attrib- 
ute of perfection. The proposition that beneficence is an "attribute of perfection" 



2 "'Sliarh al-Usiil, p. 117. 
2,12 Above, pp. 61-62, 64. 
2<I, A'. al-Majmu', p. 66. 

2114 K. Nihaya al-lqdam, p. 46. Cf. Ash'ari. Maqalal al-lslamiyin, ed. H. Ritter (Istanbul. 1929- 
1930). p. 182. 



80 



Proofs of Eternity from Nature of God 



in God cannot, he writes, be accepted a priori, since it is not an item "of necessary 
or immediate [knowledge]"; 205 and since there is no way in which it can be 
known a posteriori either, the proposition is groundless, furthermore, and more 
importantly, beneficence could not possibly be an attribute of perfection in God. 
For if beneficence were such an attribute, God's perfection would be dependent 
on the existence of his creatures, who are the expression of his beneficence, 206 
and the "superior would acquire perfection through the inferior, which is absurd." 
Seeing that beneficence is not an attribute of perfection and does not pertain to 
God's essence, the eternity of the world cannot be inferred from it. 207 

In the instances to follow the adherents of creation do not merely refuse to 
admit the eternity of the attributes from which the eternity of the world had been 
inferred. They refuse to admit that the attributes or traits are in any sense pos- 
sessed by God. Ibn Hazm recorded an argument for eternity turning on the 
proposition that God could only have produced the world insofar as he is a cause 
"by reason of himself." If God produces the world insofar as he is a cause by 
"reason of himself," the thinking went, his effect must have existed for as long 
as he has himself existed. 208 The response Ibn Hazm makes is that God docs not 
act by "reason of himself" nor, indeed, "by reason of" anything whatsoever. 
God simply acts "as he wishes" to act. Since the proposition that God acts by 
reason of himself is false, nothing can be inferred from it. 209 

Gersonides reconstructed and ascribed to Aristotle an argument for eternity 
from the premise of the deity's being the mover of the outermost sphere. The 
reconstructed reasoning was that since the deity is by his nature the mover of 
the outermost sphere, the sphere must exist as long as the deity exists. 210 Ger- 
sonides responds by denying that the deity has the trait in question. It can be 
demonstrated, he maintains, that the ultimate cause of the universe is not in fact 
the mover of the outermost sphere but an entity distinct from, and beyond the 
movers of the spheres. 2 " Inasmuch as the deity is not by his nature the mover 
of the sphere, the eternity of the sphere cannot be deduced from the eternity of 
the deity. 212 

Fakhr al-DIn al-RazI, to take a final example, knew of an argument for eternity 
from the deity's character of being an "agent." To be an agent, the argument ran, 



205 Cf. Ghazali, Tahafut al-Falasifa, I, §§13 and 33 (English translation, pp. 6 and 21), where 
"necessary" knowledge, in the sense of immediate knowledge, is contrasted with "speculative" 
knowledge. For the term "necessary knowledge," cf. J. van Ess, Die Erkenntnislehre des 'Adudaddin 
al-lci (Weisbadcn, 1966), pp. 116, 118. 

206 Cf. A. Lovejoy, The Great Chain of Being (Cambridge, Mass., 1936), pp. 54, 62. 

2m Ghaya al-Maram, p. 270. 

208 Above, p. 66. 

2m K. al-Faslfi al-Milal, I, p. 10. 

2,0 Above, p. 67. 

2ll This is Aviccnna's position; above, n. 134. 
2,2 Milhamot ha-Shem, VI, i, 4, p. 303. 



Proofs of Eternity from Nature of God 



81 



is an "eternal . . . positive attribute" added to God's essence, and it is moreover 
a "relative attribute." Since the attribute is eternal, positive, and relative, its 
presence implies the existence ol an eternal positive correlative, an eternal effect. 2 " 
Razi's response is that to be an agent is not — and could not possibly be 2 ' 4 — a 
positive attribute added to God's essence. No light, therefore, can be shed on the 
issue of eternity and creation by God's being an agent. 215 

This has been one form of response to arguments for eternity from the eternal 
divine attributes. A second form of response consists not in denying that the 
attributes arc eternal or that God possesses them, but in denying that they entail 
the eternity of the world. When answering an argument from the divine attributes 
of wisdom and knowledge, 216 'Abd al-Jabbar agrees that God's wisdom and his 
knowledge of the goodness of the world are eternal. But, 'Abd al-Jabbar insists, 
wisdom "docs not necessitate action." Therefore, although God undoubtedly 
knew from eternity that a world should exist and that he would create one, his 
knowledge did not necessitate the production of the world from eternity. 217 Ibn 
Hazm, for his part, refuses to acknowledge that God's "beneficence, wisdom, 
and power" arc at all the "cause" of God's creating the world. "There is," he 
asserts, "no cause of God's bringing things into existence"; God simply created 
the world, and nothing more can be known or said. Consequently, the attributes 
of beneficence, wisdom, and power, whatever their status might be, cannot be 
understood to have necessitated the existence of a world, and certainly not the 
existence of an eternal world. 2IS 

The argument from God's eternal beneficence is also answered by Fakhr al- 
DTn al-RazI. RazI docs not dispute the premise that God is eternally beneficent. 
He submits, however, that the attribute of eternal beneficence need not express 
itself in an eternal world, and to support his position, he employs a device already 
met in the answer to another argument for eternity. 219 Even on the assumption 
of the eternity of the world, Razt notes, God's beneficence is expressed in the 
production of myriad objects — "forms and accidents" — that "come into exis- 
tence" and arc not eternal. All parties must thus recognize that God's eternal 
attribute of beneficence is compatible with noneternal products, and no one should 

21, Abovc, pp. 66-67. 

2 14 Thc reason RazI gives is a form of the "third man" argument : If being an agent were an attribute 
added to the essence, it would be an entity of some sort dependent on the essence, and hence it would 
be a possibly existent entity (cf. below, p. 290). As possibly existent, it would stand in need of an 
agent to give it existence. The agent could only be God. But if God's character of being the agent 
that gave His attribute of being an agent existence were, then, itself an attribute added to God's 
essence, it would, in turn, also stand in need of an agent. And so on. 

2I? X. al-Arba'in, p. 51. 

2 "'Above, p. 60. 

2[7 Sharh al-Usfd, p. 1 16. 

2I8 A'. al-Fasl fl al-Milal, 1, p. 20. 

2l "Abovc, pp. 70, 71-72. 



82 



Proofs of Eternity from Nature of God 



balk at the hypothesis that God's eternal beneficence can give rise to one more 
noneternal product, namely the universe. 220 

Aquinas replies to the argument from God's "infinite" goodness by observing 
that God's goodness should be expected to express itself by producing creatures 
who "represent the divine goodness" in the most accurate possible manner. The 
relation of the creator to what he creates is surely not a relation of equals but a 
relation of the superior to the inferior, of the infinite to the finite. And the "supe- 
riority of divine goodness over what it creates is best expressed | precisely] in the 
latter's not being eternal." The eternal attribute of infinite divine goodness has 
its most appropriate expression in a noneternal, rather than eternal, product; 
divine goodness, far from entailing eternity, harmonizes better with creation. 221 
A similar response to the argument from divine wisdom and beneficence is made 
by Abravanel. 222 In an alternative and more succinct response to the argument 
that God's eternal goodness must give rise to an eternal world, Aquinas posits 
that "God's goodness does not exist ... for the sake of creatures"; consequently, 
"divine goodness would not have been idle [and thereby deficient] even if it had 
never produced a creature." 223 

Maimonides offers a reply to the argument that the eternity of the world is 
entailed by God's eternal attribute of wisdom. The proposition that God's "wis- 
dom is eternal like his essence" and the proposition that the existence of the world 
"Hows from his eternal unchanging wisdom" arc embraced by Maimonides with- 
out reservation. But, Maimonides affirms, the human mind "is completely igno- 
rant of" God's essence and wisdom. Therefore, although it is true that God's 
eternal wisdom did express itself in the existence of the universe, nothing justifies 
the conclusion that God's eternal wisdom had to express itself in an eternal 

224 

universe. 

An argument in which eternity is inferred from God's knowledge or thought 
was recorded by Aquinas and by Gersonides. In Aquinas' version, the reasoning 
was that the world must be eternal because God's knowledge needs an eternal 
world to serve eternally as its object. 225 In Gersonides' version, the starting 
premise was that God's thought, which is identical with his essence, comprises 
"the [intelligible] order of the universe." The reasoning then went: The actual 
and external order of the universe must exist whenever God and his thought, 
which comprises the intelligible order of the universe, exist; and since God and 



22a K. al-Arba'm, p. 53. The consideration that God's beneficence can express itself in noneternal 
objects would only counter an argument to the effect that eternal beneficence must express itself 
exclusively in something eternal. It would not counter the argument that eternal beneficence must 
eternally express itself in something. 

22, Sumnia contra Gentiles, II, 35(7). 

222 Mif'alot, VII, 4. The notion also appears in Leone Ebreo, Dialoghi d'Amore, p. 239. 
22i Commentary on De Caelo, I, §66. Cf. De Potentia, q. 3, art. 17 (ad 14). 
224 Guide, II, 18(3). The notion also appears in Leone Ebreo, Dialoghi d'Amore, p. 239. 
"'Above, p. 63. 



Proofs of Eternity from Nature of God 



83 



his thought arc eternal, the actual universe must be eternal. ' Aquinas and 
Gersonides make virtually identical replies to the argument. God's knowledge 
and thought, they stress, is in no sense dependent on the world, being, instead, 
the cause of the world and prior thereto. Consequently, in Aquinas' words: "God 
can have knowledge even should the object of his knowledge not exist." 227 And 
in Gersonides' words: The "[intelligible] order may exist even if the existent 
things, whose existence can flow from that [intelligible] order, do not exist." 228 
God could, that is to say, have eternal knowledge of what the universe would 
eventually be after it came into existence. 229 

Aquinas recorded two additional arguments for eternity from the eternal attri- 
butes of God and he finds that those arguments also fail because the attributes 
they focus on, although belonging eternally to God, do not imply eternity. One 
argument went: Since God is a sufficient cause from all eternity, and since when- 
ever a "sufficient cause is given, its effect is given," God's effect must exist from 
all eternity. 230 In his reply, Aquinas has recourse to the thesis that the creator is 
a voluntary agent. The "proper effect" of a voluntary cause, he writes, is an 
effect that comes about as "the will wills," that conforms to the "intention" of 
the agent. In other words, the proper effect is nothing other than what is willed, 
and if what a voluntary cause wills should be a noneternal product, a noneternal 
product and nothing else is the appropriate effect of the cause. An eternal, but 
voluntary, sufficient cause could produce a noneternal world no less appropriately 
than an eternal world. 23 ' A similar response to the same argument for eternity 
is offered by Bonaventurc. 232 

There was, finally, the argument that the world must be eternal because God 
is eternally the "Lord." 233 Aquinas responds by explaining that God is the Lord 
in the sense of having the "power of governing" and not necessarily in the sense 
of actually governing. 234 God therefore can be the Lord eternally although the 
universe does not exist eternally. 



227 De Potentia, q. 3, art. 17 (ad 19). Aquinas refers to Aristotle's characterization of certain things 
as existing "together by nature"; sec Categories 7, 7b, 15. Things existing together by nature are 
things such that if cither one of the pair docs not exist, the other cannot exist. God's knowledge and 
the object of his knowledge plainly do not meet the description; for God's knowledge precedes the 
world and is its cause, whereas the converse is not true. And that is the reason "God can have 
knowledge even should the object of his knowledge not exist." 

""Milhamot ha-Shem, VI, i, 4. pp. 303-304. 

22il Cf. 'Abd al-Jabbar's response to the argument that creation would entail change in the creator's 
knowledge; above, pp. 78-79. 
2, "Abovc, p. 66. 

2>, Snmina contra Gentiles, 11. 35 (ad 3). Cf. Summa Theologiae, 1. 46, art. I . ad 9; De Potentia, 
q. 3. art. 17 (ad 6). 

232 Commentary on II Sentences, d. 1 , p. 1 , a. I . q. 2. 
"'Above, p. 64. 

2M De Potentia, q. 3, art. 17 (ad 21). 



84 



Proofs of Eternity from Nature of God 



The third form of response to arguments from the eternity of God's attributes 
consists in showing that it would be utterly impossible for God's attributes to 
give rise to an eternal universe. Two procedures are in evidence: One rests on 
the intrinsic impossibility of an eternal universe, without qualification; the other, 
on the intrinsic impossibility specifically of an eternal universe that has a cause 
of its existence. 

The former procedure is employed by John Philoponus, who builds on his 
proof of creation. Proclus had argued that the eternity of the world is implied by 
the deity's "goodness," and Philoponus counters — as he and others had done 
when answering some of the previous arguments for eternity 235 — with the truism 
that God can produce only what is possible. Since it can be demonstrated defin- 
itively, so Philoponus understands, that an eternal world is absolutely impossible, 
even God could not produce such a world. The most that God could produce, 
the most his goodness could express itself in, is a created world. If an eternal 
world is impossible, the eternity of God's goodness obviously cannot entail the 
eternity of the world. 236 The same position is taken in the Middle Ages by Amidl. 
In the course of an elaborate rejoinder to the argument from God's "beneficence," 
Amidl remarks that God's failure to produce the world from eternity betokens no 
lack of beneficence on his part; for the eternity of the world is demonstrably 
impossible, 237 and "there is no defect in failing to produce what is impossible." 238 

Ibn Hazm employs the other procedure, the procedure resting not on the 
unqualified impossibility of the world's being eternal, but on the impossibility 
that the world should be eternal and nevertheless have a cause of its existence. 
To have a cause of existence and to be eternal are, Ibn Hazm explains, mutually 
exclusive. For "an effect is what passes from nonexistence to existence . . . and 
thus is tantamount to that which comes into existence. The meaning oilhat which 
comes into existence is that which does not exist and subsequently exists, . . . 
something quite different from that which is eternal.'" To state that the world has 
a cause and is an effect, yet is eternal, is accordingly to enunciate a "total 
absurdity," an absolute impossibility. And what is absolutely impossible lies out- 
side even the deity's control. Now the argument from "divine beneficence, wis- 
dom, and power" does acknowledge — in fact presupposes — that the world has 
a cause, that it is an effect, and hence that it has been brought into existence. 
Given this acknowledgment, the argument, far from establishing the eternity of 
the world, falls into a blatant contradiction by concluding that the world is eternal. 
The argument is therefore wholly invalid. 239 Ibn Hazm's response to the argument 
from divine beneficence, wisdom, and power, as outlined here, would apply, of 



Above, pp. 69, 78. 
' De Aeternitate , p. 8. 
'Cf. below, p. 191. 
'Ghaya al-Maram, p. 270. 
'K. al-Fasl ft ai-Milal, 1, p. 20. 



Proofs of Eternity from Nature of God 



85 



course, to any argument for eternity from God's eternal attributes, and indeed to 
any argument for eternity which presupposes a cause of the existence of the 
universe. 

ShahrastanI makes virtually the same response to the argument from God's 
beneficence as Ibn Hazm. 240 

3. Summary 

The first argument for eternity from the nature of the cause of the universe 
reasoned that the creator could not have selected a given moment for creation in 
preference to any other moment. The adherents of creation respond that although 
all moments before creation were equal, and no explanation can be provided for 
the creator's having chosen one in preference to another, a particular moment 
could have been selected arbitrarily by God's eternal will. Subordinate consid- 
erations are that there was no time before creation, hence no moment was strictly 
selected out of an infinite time; that deist adherents of eternity are in the same 
predicament as the adherents of creation, since they for their part must recognize 
arbitrary determinations by God in the spatial realm; that adherents of eternity 
who trace events occurring in the universe back to the divine will must recognize 
the ability of an eternal will — albeit through intermediaries and within the frame- 
work of an already existing universe — to make determinations in the temporal 
realm; that once creation is demonstrated and the world is known to have come 
into existence at some moment, it is pointless to ask why creation took place at 
a certain moment rather than at another. 

The second argument for eternity from the cause of the universe reasoned that 
an ultimate cause must be unchangeable, whereas creation would imply a tran- 
sition from inaction to action, and hence change, in the cause ultimately respon- 
sible for bringing the universe into existence. The adherents of creation usually 
respond by inverting the thinking; they assert that since the cause of the universe 
must be unchangeable, creation would not, in his case, imply change. An alter- 
native response is that the creator is eternally and unchangeably in the state of 
bringing the world into existence, but since an eternal world is impossible, the 
world could come into existence no sooner than in the finite past. 

The third argument was that God's beneficence and other divine attributes 
express themselves in the existence of the world; and since they arc eternal, the 
existence of the world, which is their expression, must likewise be eternal. The 
adherents of creation respond either by denying that God eternally has the attri- 
butes in question; by denying that the attributes need express themselves in an 
eternal world; or by showing that it would be intrinsically impossible for the 
attributes to express themselves in an eternal world. 



'K. Nihaya al-Iqdam, p. 47. 



IV 



John Philoponus' Proofs of Creation 
and Their Entry into 
Medieval Arabic Philosophy 



1. Philoponus' proofs of creation 

John Philoponus carefully distinguished the negative task of refuting the argu- 
ments advanced by proponents of eternity from the positive task of proving 
creation. 1 To accomplish the latter task, to prove creation, he drew up two sets 
of proofs, one built around the impossibility of an infinite number, the other based 
on the principle that a finite body can contain only finite power. Both sets of 
proofs are mentioned in Philoponus' surviving works, but only in passing. 2 Their 
systematic development was undertaken by Philoponus in at least two works no 
longer extant: the Contra Aristotelem, and another brief work which was some- 
how connected to it. 3 The Contra Aristotelem and the related work have been 
partially preserved in Simplicius' commentaries on Aristotle, 4 and either they or 
Simplicius' excerpts from them were, as will appear, accessible to the medieval 
Arabic philosophers. 

Philoponus' first set of proofs for creation comprises three arguments, each 
giving a different reason why an infinite series of past events is impossible. The 
conclusion in each instance is that the world cannot have existed for an infinite 
time but must have had a beginning. 

'John Philoponus, De Aeiernitate Mundi Contra Proclum, ed. H. Rabe (Leipzig, 1899), p. 9; 
Simplicius, Commentary on Physics, ed. H. Dicls, Commentaria in Aristotelem Graeca, Vol. X 
(Berlin, 1895), p. 1 178, top; S. Pines, "An Arabic Summary of a Lost Work of John Philoponus," 
Israel Oriental Studies, II (1972), 322. 

2 Philoponus, De Aeiernitate, pp. 8-9, 325. 

'Regarding that brief work and its connection with the Contra Aristotelem, see H. Davidson, 
"John Philoponus as a Source of Medieval Islamic and Jewish Proofs of Creation," Journal of the 
Amerian Oriental Society, LXXXIX (1969), 358-359. 

'Simplicius, Commentary on Physics, pp. 1171 ff.; Commentary on De Caelo, ed. I. Hciberg, 
Commentaria in Aristotelem Graeca, Vol. VII (Berlin, 1894), pp. 28 ff. 



86 



Philoponus' Proofs of Creation 



87 



(</./') The first of the three arguments pointedly draws from Aristotle. In the 
De Generatione el Corruptione, Philoponus recalls, Aristotle rejected the thesis 
that the transformation of the physical elements into each other "goes to infinity 
in a straight line," that is, the thesis that earth, for example, becomes water, water 
becomes air, air becomes lire, and lire, rather than reverting — circularly, as it 
were — to air is transformed into a fifth type of clement, and so on ad infinitum. 
One of Aristotle's objections to the thesis in question was that at least some of 
the hypothetical infinite elements could come into existence only after an infinite 
number of transformations, whereas — as Philoponus spells out Aristotle's 
meaning 5 — "the infinite cannot be traversed," so that the point never could be 
reached where those elements actually did come into existence. Consequently, at 
least some of the elements in the supposed infinite scries could never exist, and 
if part of the series could never exist, the scries itself could not be an actuality.*' 

Aristotle's reasoning is subtly turned by Philoponus against the assumption of 
the eternity of the world. Philoponus does not consider the transformations occur- 
ring "in a straight line" through an infinite series of different types of physical 
elements. Instead he takes up the no less infinite series of transformations that, 
on the assumption of eternity, necessarily have taken place circularly among the 
four recognized types of elements, and linearly among individual elemental par- 
ticles. Philoponus reasons: The existence of a given particle of the clement lire 
must have been preceded by the generation of that fire from a particle of air, the 
existence of the air must have been preceded by its generation from, let us say, 
water, the water from cither earth or air, and so on. In an eternal world, these 
transformations would constitute an infinite scries. Now it is evident that an 
infinite number can neither "actually" exist nor be traversed. Therefore, in an 
eternal world, the infinite scries of transformations leading up to the generation 
of a given particle of fire could never be completed, and the particle known to 
exist at the present moment could never have come into existence. "The same 
argument," Philoponus adds, "can be applied to other particular motions" as 
well, that is, to the various scries of transformations leading up to the emergence 
of whatever individuals exist in the world today. 7 In each instance, an infinite 
series of transformations would, on the assumption of eternity, have to be trav- 
ersed in order that what exists today might emerge. 



5 Cf. Philoponus. Commentary on De Generatione, cd. H. Vitclli, Commentaria in Aristotelem 
Graeca, Vol. XIV/2 (Berlin, 1897). p. 254. 

''Cf. Aristotle, De Generatione et Corruptione II. 5. 332b. 12 — 333a, 12. The concept of a 
circular scries appears in II, II, 338a. 4 ff. The principle that an inlinite cannot be traversed is not 
explicitly stated by Aristotle there, but is adduced by him elsewhere in connection with both the 
spatial and nonspatial realms. Cf. H. Bonitz, Index Aristotelicus (Berlin, 1870), p. 74b. lines 30- 
34. and especially the references to Posterior Analytics I. 22. 83b, 6; and Physics, 204b. 8-10; 
263b. 4; 265a, 19. 

'Simplicius, Commentary on Physics, pp. 1 178-1 179. 



88 



Philoponus' Proofs of Creation 



When Philoponus states the argument concisely, he rests it on the general rule 
that "the infinite cannot be traversed"; and the inference drawn is that if the 
"number of individuals going up (avw) [into the past]" were infinite, the process 
of generation could never have been traversed and "come down to each of us." 8 
When Philoponus presented the argument more fully, so Simplicius informs us, 
he employed the same general rule but supplemented it with an additional rule. 
The additional rule affirms that if something must be preceded by something else 
in order to be generated, then "the former will not be generated unless the latter 
is generated prior thereto." In other words, if the prior existence of y is a condition 
for the coming into existence of x, then x obviously will never come into existence 
unless y has already been in existence. The inference now drawn is that "if for 
the generation of a given thing there must first exist an infinite number of things 
that are generated from one another, then the given thing cannot be generated." 9 
If x cannot come into existence unless an infinite series of things has already 
come into existence — a condition that cannot conceivably be fulfilled — x will 
never come into existence. Students of Kant will observe the similarity to the 
proof of the thesis of the first antinomy. 

The upshot of Philoponus' first proof of creation is that transformations in the 
sublunar world cannot be "conceived to precede one another infinitely," but must 
have a beginning; and since the translunar world is inextricably connected with 
the sublunar world, 10 it too must have a beginning." 

(a.ii) Philoponus' second argument for creation from the impossibility of an 
infinite number employs a principle which although apparently not strictly Aris- 
totelian, 12 does appear in the Peripatetic tradition. The principle affirms that since 
an infinite cannot conceivably be exceeded, one infinite cannot be greater than 
another. 13 Philoponus applies the principle thus: "Since motions yet to be gen- 
erated, when added to those already generated, increase their number, and since, 
moreover, the infinite cannot be increased, it follows that the motions already 
accumulated cannot be infinite." 14 That is to say, each new movement in the 
sublunar world and each new revolution of the celestial spheres add to the number 
that has gone before. If the number that has gone before should be infinite, each 
new movement would add to an infinite number. But the infinite cannot be increased. 

"Philoponus, De Aeternilale, pp. 10-11; cf. his Commentary on Physics, cd. H. Vitclli, Com- 
mentaria in Aristotelem Graeca, Vol. XVI (Berlin, 1887), pp. 428-429. 

'Simplicius, Commentary on Physics, p. 1 178. On p. 1 179, Philoponus is reported (o have spoken 
of "going back up (otvoSos)" through the infinite. 

10 Cf. Aristotle, De Caelo II, 3; Philoponus, as reported by Simplicius, Commentary on Physics, 
p. 1179. 

"Simplicius, Commentary on Physics, p. 1 179. 

12 In fact, Physics III, 6, 207a, 1 ff., seems to state the contrary. 

"Alexander of Aphrodisias, Aporiai, ed. I Bruns, Commentaria in Aristotelem Graeca, Supple- 
mentary Vol. II/2 (Berlin, 1892), III, 12, p. 103. Alexander observes that the assumption of an 
infinite multiplied infinitely would be the highest absurdity. 

' 4 Simplicius, Commentary on Physics, p. 1179. Cf. Philoponus, De Aeternitate, p. 11. 



Philoponus' Proofs of Creation 



89 



The number of past motions must therefore be finite, and the universe can have 
existed for only a finite time. 

(a. Hi) Philoponus' third argument is not independent of the previous two. It is 
a variation of the second, repeating, for good measure, the central thought of the 
first. Whereas the second argument rested on the principle that an infinite cannot 
be increased, the third rests on the principle that an infinite cannot be multiplied; 
multiplication, though, is simply a type of addition. To reinforce the argumen- 
tation, Philoponus again refers to the rule that an infinite cannot be traversed. 

Philoponus' attention is here restricted to the heavens, and his reasoning is 
reported thus by Simplicius: "If . . . the motion of the heavens has no beginning, 
the sphere of the planet Saturn has performed infinite revolutions, the sphere of 
the planet Jupiter almost three times as many, the revolutions of the sun will be 
thirty times those of Saturn, the revolutions of the moon will be 360 times as 
many, and the revolutions of the fixed sphere [which revolves once each twenty- 
four hours] will be more than 10,000 times as many. Considering that the infinite 
cannot be traversed even once, is it not beyond all absurdity to suppose the infinite 
multiplied by 10,000, nay multiplied infinitely? It necessarily follows that the 
circular motion of the heavens had a beginning." 15 Aristotelians understood cir- 
cular motion to be an essential expression of the nature of the spheres. As long 
as the spheres exist they would have to perform such motion and could not exist 
in a state of rest. 16 Once Philoponus has established that the motions of the 
heavens had a beginning, he can therefore conclude as well that "the heavens 
themselves also had a beginning of their existence." 17 

(/?./) Philoponus' second set of proofs for creation rests on the Aristotelian 
principle that a finite body can contain only finite power. The principle has a 
peculiar history. 18 Its career began in Aristotle's proof of the existence of a first 
mover, where it served as an essential premise in establishing that the first mover 
is incorporeal. Aristotle reasoned that infinite power is needed to sustain the 
motion of the universe for an infinite time; that all bodies are finite; and that a 
finite body can contain only finite power. The conclusion he drew was that the 
first cause of the motion of the universe cannot be a body, but must be an incor- 
poreal being. 19 Proclus subsequently took up the Aristotelian principle that a 
finite body can contain only finite power and arrived at a more far-reaching result 
than Aristotle. He converted Aristotle's argument that the motion of the universe 
must depend upon an incorporeal cause into an argument showing that the very 
existence of the physical universe must depend upon an incorporeal cause. Proclus 
reasoned that infinite power is needed to sustain not only eternal motion, but 

"Simplicius, Commentary on Physics, p. 1 179. Philoponus, De Aeternitate, p. 1 1 . Cf. above, 
n. 13. 

'"Aristotle, De Caelo I, 2. 

"Simplicius, Commentary on Physics, p. 1 179. 

'"Sec Appendix A. 

"Aristotle, Physics VIII, 10; below, p. 244. 



( 



90 Philoponus' Proofs of Creation 

eternal existence as well; and since the finite physical universe can contain only 

finite power, the physical universe must depend upon an incorporeal being outside 
[ itself for its very continued existence. 20 Philoponus knew of the arguments of 

i Aristotle 21 and of Proclus, 22 and they led him to a proof of creation. Since the 

corporeal universe contains only finite power, the universe, such will be Philo- 
C ponus' argument, is not merely incapable of sustaining its own eternal motion 

or eternal existence; it is incapable of existing eternally no matter what the cause 

sustaining its existence might be. As will be seen in later chapters, Philoponus' 
( adaptation of the argument from the finite power of finite bodies would elicit a 

response from Averroes, 23 and Averroes' response would, in turn, elicit from 
^ Crescas a critique of the original Aristotelian proof of the incorporeality of the 

( first mover. 24 

Here we are interested in Philoponus' proof. Philoponus is reported by Sim- 
' plicius to have contended: Inasmuch as "the body of the heavens and of the 

( universe is finite, it contains [only] finite power. And what contains finite power 

. . . [is] destructible." 25 That is to say, since the corporeal universe contains only 
( finite power, it is, considered in itself, incapable of existing through infinite future 

( time and consequently is liable to destruction. 26 The proposition that the corporeal 

universe is liable to destruction will not be invalidated even by the assumption 
( that the universe is maintained in existence by a "transcendent" power as, for 

; example, the "will of God." 27 For even should such an assumption be made, the 



universe will remain incapable of existing infinitely into the future, insofar as it 
' is considered in itself; the universe, considered in itself, will remain liable to 

r destruction. Stated otherwise, the universe will still have the "logos (nature or 

ground) of destruction." 28 Now a well known and widely employed 29 Aristotelian 
( principle affirms that over an infinite time every possibility must eventually be 

I realized. 30 Accordingly, Philoponus concluded, the logos of destruction in the 

universe must "sometime come to actuality" and the universe must some day 
( actually be destroyed. 31 



(" 

( 20 Cf. below, pp. 281-282. 

21 Simplicius, Commentary on Physics, p. 1327; Philoponus, De Aeternitate, p. 238. 
( "Philoponus, De Aeternitate , pp. 238-240. 

"Below, pp. 323 ff. 24 Below, pp. 264-265. 

( 25 Simplicius, Commentary on Physics, p. 1327. 

26 Cf. Philoponus, De Aeternitate, p. 235. 
| 27 Simplicius, Commentary on Physics, pp. 1330, 1331, 1333. 

28 lbid., pp. 1331, 1333 . 29 Cf. below, pp. 320, 381. 

( 30 Aristo(le, Physics III, 4, 203b, 30; Metaphysics IX, 4, 1047b, 4-5; 8, 1050b, 8-15. Cf. 

J. Hintikka, Time and Necessity (Oxford, 1973). pp. 95-96, 103-105, 107. 

■"Simplicius, Commentary on Physics, p. 1333. Cf. Aristotle, De Caelo 1, 12, 281b. 20-25 and 
I 283a, 24-25; Metaphysics XIV, 2; Alexander, Aporiai, 1, 18; and the passage translated by Pines, 

"An Arabic Summary of a Lost Work of Philoponus," pp. 324-325. In Simplicius, Commentary on 
\j Physics, pp. 1331, 1333, and Philoponus, De Aeternitate, pp. 241-242, Philoponus seems to coun- 

tenance the possibility of the future eternity of the world. 



Philoponus' Proofs of Creation 



91 



That is the extent of Simplicius' report, and it falls short of an explicit proof 
of creation. Another work by Philoponus makes clear, however, how Philoponus 
passed from the dcstructibility, and actual future destruction, of the world to 
creation. He employed yet another well known Aristotelian principle, one that 
happens to rest on the principle/just cited, according to which every possibility 
must eventually be realized. The other principle affirms that "whatever is des- 
tructible must be generated." 32 If something is destructible, if it has the possibility 
of not existing, that possibility must have been realized at some point in the 
infinite past just as it must be realized at some point in the infinite future. Just as 
a destructible universe must eventually become nonexistent in the future, so too 
must a destructible universe have been nonexistent in the past. Our universe, 
therefore, cannot have existed forever, but must have come into existence. 33 

Philoponus' reasoning, then, was this: Since a finite body can contain only 
finite power, the corporeal universe has the logos of destruction. Since, moreover, 
every possibility is eventually realized, the universe must eventually undergo 
destruction. And since whatever is destructible is generated, the universe must 
have been generated. The Aristotelian principle that a finite body can contain 
only finite power, taken together with two other Aristotelian principles, leads to 
the highly un-Aristotelian conclusion that the world can neither exist forever in 
the future nor have existed forever in the past. Apparently, Philoponus understood 
that he was demonstrating not merely the creation of the universe in its present 
form but the creation of the matter of the universe ex niliilo; for his contention 
is that the very matter of the universe, being finite, lacks the power to sustain 
itself in existence for an infinite time. 

Philoponus was not content simply to cite the authority of Aristotle for the 
principle that a finite body can contain only finite power. He proceeded to offer 
a set of five or six 34 auxiliary arguments, all ostensibly intended to support the 
principle. In fact, the auxiliary arguments are not uniform. Some of them do 
support the proposition that the heavens, or the entire universe, can contain only 
finite power; the proof for creation is then to be completed through the consid- 
eration that what contains finite power is destructible, hence also generated. 
Inexplicably, however, some of the auxiliary arguments 35 move in the other 
direction. They start by establishing that the corporeal universe is destructible 
and infer therefrom that the universe can contain only finite power. The reader 
is left either again to draw the further inference, now quite redundant, that what 

"Aristotle, De Caelo I, 12, 282b, 2; ibid., 281b, 26— 282a, 1. 
"Philoponus, De Aeternitate, pp. 225, 230. 235, 240, 241. 

"Simplicius, Commentary on Physics, enumerates only four auxiliary arguments. But on p. 1332, 
between the "third" and "fourth." he quotes an additional auxiliary argument not included in the 
enumeration; and the auxiliary argument that he enumerates, p. 1335, as the "fourth," contains two 
separate considerations. In the article referred to above, n. 18, 1 outline all the arguments except the 
one designated by Simplicius as the "second." That argument is dealt with here, below (b.iii). 

15 See below, (b.iv). 



92 



Philoponus' Proofs of Creation 



has finite power is destructible, and hence is generated; or else to derive the 
generation of the universe directly from its finite power. One of Philoponus' 
auxiliary arguments even drops the issue of finite power altogether. 36 

For our purposes, the first three of Philoponus' auxiliary arguments are of 
interest. 

(b.ii) In the "first" of his auxiliary arguments, Philoponus is reported to have 
reasoned that "the heavens are [composed) of matter and form; what is (com- 
posed) of matter requires the matter for its existence; what requires something is 
not self-sufficient; what is not self-sufficient is not infinitely powerful. From all 
this he concludes that the heavens, considering their own nature, are not infinitely 
powerful and therefore are destructible." 37 It remains for us to add that what is 
destructible must be generated, and consequently the heavens and the rest of the 
corporeal universe cannot — at least in their present form — have existed for all 
eternity but must have come into existence. 

(b.iii) Philoponus' "second" auxiliary argument, as reported by Simplicius, 
drops the issue of finiteness of power. Philoponus contends: "The essence of 
matter consists in its being suited to receive all forms. It docs not possess that 
power in vain; the same matter cannot admit several forms at once; nor can matter 
retain any form eternally insofar as its own nature is considered." Since matter 
is by nature such that it does not retain any form permanently, "nothing |com- 
poscd] of matter and form will, considering its matter, be indestructible." The 
corporeal universe is thus destructible. 38 The reader is left to supply the principle 
that everything destructible is also generated and to draw the conclusion that the 
corporeal universe in its present state — clearly not the matter of the universe — 
must have come into existence. 

(b.iv) The "third" of the auxiliary arguments opens, like the first, with the 
contention that the heavens are "composed ... of substratum . . . and form," 
the latter here being specified as "solar or lunar form" and the forms of the other 
heavenly bodies. Anyone who would exclude the distinction of matter and form 
from the heavens would, Philoponus adds, still have to acknowledge the presence 
in the heavens of "extension in three dimensions," which too is a mode of com- 
position. Philoponus does not hereupon reason, as he had done in his first aux- 
iliary argument, that composition implies the absence of self-sufficiency, hence 
the absence of infinite power. Instead, he pursues a line of thought going back 
to Plato's Phaedo, where the proposition had been laid down that the fact of 
something's being composite implies its being subject to decomposition. 39 What- 



36 Beiow, (b.iii). 

""Simplicius. Commentary on Physics, p. 1329. The train of reasoning is to a large extent 
borrowed from Proclus, Elements of Theology, ed. and trans. E. Dodds (Oxford, 1963), §127. 
38 Simplicius, Commentary on Physics, p. 1329. 

39 Phaedo, 78C. Aristotle, Metaphysics XIV, 2, 1088b, 14-28, contends that what is composite 
has the potentiality of not existing, and what has the potentiality of not existing is not eternal. 



Philoponus' Proofs of Creation 



93 



ever has the "logos (nature or ground) of composition" has, Philoponus explains, 
the "logos of dissolution" and the "logos of destruction." And what has the "logos 
of destruction is not infinitely powerful." The heavens, therefore, are not infinitely 
powerful. 40 Here, as will be observed, is an instance of Philoponus' moving not 
from the finiteness of the power of the heavens to their destructibility, but vice 
versa. The argument must be completed either by reasoning, redundantly, back 
again from the finite power of the heavens to their destructibility and then adduc- 
ing the principle that everything destructible is generated; or else by inferring 
the generation of the heavens without returning to their destructibility. The infer- 
ence might, for example, be that what has finite power cannot maintain itself 
indefinitely, hence contains the possibility of not existing, and hence cannot have 
existed forever. 

To summarize: Philoponus formulated two sets of proofs of creation, both of 
which employ Aristotelian principles to draw the un-Aristotelian conclusion that 
the world is not eternal but had a beginning. Each set of proofs rests on the 
impossibility of an infinite of one sort or another. The first set argues in three 
different ways that an infinite number of events cannot conceivably have preceded 
the present moment. The second set argues that a finite body cannot contain 
infinite power, that what contains only finite power has the possibility of not 
existing, and what has the possibility of not existing cannot have existed forever. 
The second set comprises a general statement of the proof and supporting arguments. 

The most important source of the first of the two sets of proofs is Philoponus' 
Contra Aristotelem. The Contra Aristotelem, which is no longer extant and which 
is accessible today solely through Simplicius' excerpts, was available in some 
form to the medieval Arabs. 41 It is listed by the Arabic bibliographers, 42 a fact 
that, by itself, does not necessarily mean the book was translated into Arabic. 
But, in addition, passages from the book, cited in Philoponus' name, have been 
discovered in the writings of Alfarabi, 43 Sijistanl 44 (9 1 2 - 985), and Avicenna, 45 
although unfortunately none of the passages that have been discovered is from 
the section containing the proofs for creation. Philoponus' second set of argu- 
ments, from the finite power of the corporeal universe, was developed in a sepa- 
rate work that might possibly have been written as an appendix to the Contra 



""Simplicius, Commentary on Physics, p. 1331. 

■"Simplicius is probably not the source, since his commentaries on the Physics and l)e Caelo, 
which preserve the excerpts from Philoponus, are not mentioned by the medieval Arabic bibliogra- 
phers. (For the Arabic bibliographers' knowledge of Simplicius, I am relying on M. Stcinschneider, 
Die Arahischen Vehersetzungen aus dem Griechischen [rcpr. Graz, I960].) 

42 Cf. Ibn al-Nadim, K. al-Fihrist (Leipzig, 1871), I, 254, and. for al-Qifti and Ibn Abi Usaybi'a, 
M. Stcinschneider, Al-Farahi (St. Petersburg, 1869). pp. 162, 220-224. 

41 M. Mahdi, "Alfarabi against Philoponus," Journal of Near Eastern Studies , XXVI (1967), 236. 

44 J. Kraemcr, "A Lost Passage from Philoponus' Contra Aristotelem, in Arabic Translation," 
Journal of the American Oriental Society, LXXXV (1965), p. 319, n. 4, and p. 325. 

45 Ibid., p. 324, n. 27. 



94 



Philoponus' Proofs of Creation 



Aristotelem. This other work was also available to the medieval Arabs. The 
Arabic bibliographers list, among the works of Philoponus, A Single Treatise 
Showing that Every Body is Finite and has Finite Power. 47 Again, that is not 
necessarily evidence that the work was translated into Arabic; but in addition, 
Yahya ibn 'AdI (892-973) knew of a treatise in which Philoponus set forth the 
proof from finite power, 48 and the main thesis of the proof is cited in Philoponus' 
name by Ibn Suwar 49 in the tenth century, and by Averroes. 50 Further evidence 
that both sets of proofs were current in the Middle Ages is furnished by a short, 
recently discovered Arabic text, which styles itself a summary of three "treatises" 
of Philoponus. In each of the three treatises Philoponus is represented as having 
given a different proof of creation. 51 One of the three proofs turns out to be a 
statement of the first of Philoponus' arguments from the impossibility of an 
infinite number, some of the wording even being identical with the version report- 
ed by Simplicius; 52 a second of the three proofs consists in a general statement 
of the proof from the finite power of finite bodies; 53 whereas the third proof 
is probably attributed to Philoponus erroneously. 54 

There is thus ample evidence that Philoponus' proofs were accessible to readers 
of Arabic in the Middle Ages. As will appear in the remainder of the present 
chapter and in the following chapter, Philoponus became a most important source 
for medieval proofs of creation. 



See Davidson, "John Philoponus as a Source of Medieval Islamic and Jewish Proofs of Crea- 
tion," pp. 358-359. 

47 Kefcrenccs above, n. 42. 

4 "Cf. S. Pines, "A Tenth Century Philosophical Correspondence," Proceedings oj the American 
Academy for Jewish Research. XXIV (1955). 115. 

49 Cf. the brief treatise by Ibn Suwar published in A. Badawi, ed., Neoplatonici apud Arahes 
(Cairo, 1955), p. 246; French translation in B. Lcwin, "La Notion dc muhdat dans Ic kalam ct dans 
la philosophic," Oriemalia Snenica. Ill (1954), 91 . 

w Averroes, Middle Commentary on Physics (Oxford, Bodleian Library, Hebrew MS. Neubaucr 
1380 = Hunt. 79), VIII, vi, 2; Long Commentary on Metaphysics (Tafsir ma ba'd al-Tahi'a, ed. 
M. Bouygcs [Beirut, 1938 — 1948 j), XII, comm. 41; Stcinschncidcr, Al-l-arabi, p. 123, citations 2- 
5. Cf. Davidson, "John Philoponus as a Source of Medieval Islamic and Jewish Proofs of Creation," 
p. 361, n. 41. 

5l Pincs, "An Arabic Summary of a Lost Work of Philoponus," pp. 320-352. 

52 Ibid., pp. 330-336. Compare the three "principles" in Pines' translation of the Arabic text 
with the three axioms listed by Simplicius, Commentary on Physics, p. 1 178. In the Arabic text, the 
key consideration that an infinite cannot be traversed is omitted, and the argument is made to rest on 
the impossibility of ever counting an actual infinite number. 

"Pines, "An Arabic Summary of a Lost Work of Philoponus," pp. 323-325. 

54 Ibid., pp. 325-329. The argument is that the lives of individuals mark off the time continuum 
into segments, each of which is finite; and finite segments of time, no matter how many there might 
be, could not join together to constitute infinite time. By contrast. Philoponus, De Aeternitalc, p. 9, 
takes the common sense position that an infinite number of past time segments, each of which was 
finite, would indeed constitute an infinite past time. 



Philoponus' Proofs of Creation 



95 



2. Saadia and Philoponus 

The Jewish philosopher Saadia (882-942) never mentions Philoponus by name 
nor quotes him directly, but Saaclia's discussion reveals that he drew from the 
two sets of proofs of creation formulated by Philoponus. 

Saadia's aim was to treat the problem of creation in all its aspects. He begins 
by offering four proofs of the generation (hadath) of the world; he demonstrates 
that the world not only had a beginning but was created ex nihilo; and he reviews 
various alternatives to the doctrine of creation ex nihilo, one of which is the 
theory that the eternal heavens arc the cause of the universe. From the entire 
discussion, seven items are of interest to us: Saadia's four proofs of the "gener- 
ation" of the world, one of several demonstrations designed to establish that the 
world was created ex nihilo, and two of his refutations — "the third" and "the 
fourth" — of the thesis that the heavens arc eternal. Five of the seven items can 
be shown unquestionably to be derived from Philoponus, and the other two can 
be traced to Philoponus with some plausibility. 

The comparison with Philoponus will be facilitated by rearranging the seven 
items from Saadia. The order in which 1 shall take them up will be: Saadia's 
fourth proof of creation, 'the argument for creation ex nihilo, the two refutations 
of the eternity of the heavens, and finally, Saadia's (irst, second, and third proofs 
of creation. The following table lists the seven items from Saadia as I have 
rearranged them, together with the corresponding arguments from Philoponus. 



Saadia 

(i) Fourth proof of creation 

(ii) An argument for creation ex 
nihilo 

(iii) A refutation of the eternity 
of the heavens 

(iv) Another refutation of the 
eternity of the heavens 

(v) First proof of creation 

(vi) Second proof of creation 

(vii) Third proof of creation 



Philoponus 

(a.i) The infinite is not traversable 
(a.i) The infinite is not traversable 

(a.ii) The infinite cannot be added 
to 

(a. iii) The infinite cannot be 

multiplied 
(b.i) Proof from the finite power 

of finite bodies 
(b.ii) Auxiliary argument from 

composition 
(b.iii) Auxiliary argument from the 

succession of forms over 

matter 



(/') Saaclia's fourth proof of the "generation" of the world carries the title "from 
time" and runs thus: Assuming that time is eternal, should a person "attempt 
mentally to ascend in time from the present moment, he would be unable to do 



96 



Philoponus' Proofs of Creation 



so, since . . . thought cannot travel up across the infinite and traverse it. The 
same reason would prevent existence from traveling down through time and 
traversing it so as to reach us. But if existence had not reached us, we should not 
exist. . . . Inasmuch as I find that I do exist, I know that existence has traversed 
time, . . . and . . . time is finite. ..." As a sort of appendix to the proof Saadia 
mentions that his position "on future time" agrees with his position "on past 
time." 55 He means that future time is, in his opinion, finite, and perhaps also that 
the finiteness of future time can be demonstrated. If he is indeed alluding to a 
demonstration of the finiteness of future time, the argument would presumably 
be that no moment in an infinitely distant future could be reached; for existence 
could not conceivably travel across and traverse infinite future time so as to reach 
whatever might be thought to exist at the infinitely distant moment. 

The proof of creation just quoted from Saadia unquestionably derives from 
Philoponus' first proof of creation from the impossibility of an infinite number. 
The key to Philoponus" proof was the rule that an infinite cannot be traversed. 56 
And the same is now the key to Saadia's proof. Characteristic details from Phil- 
oponus are also echoed in Saadia, namely the impossibility of things' existing 
today if they had to be preceded by an infinite past; the impossibility "mentally" 
(Saadia) to "conceive" (Philoponus) 57 an infinite's extending back into the past; 
and the comparison between going back "up" through the infinite past, on the 
one hand, and coming "down to each of us" (Philoponus) 5 " or "traveling down 
... to reach us" (Saadia), on the other. 

Saadia's proof docs differ from Philoponus' in a significant respect. Philoponus 
considered transformations whereas Saadia considers time, and therefore Philo- 
ponus' infinite is an infinite series, whereas Saadia's is an infinite continuum. 
The shift from the former infinite to the latter permits a certain simplification in 
the argument. Philoponus' reasoning established a beginning for the existence of 
the sublunar region, where everything undergoes transformation, but not for the 
celestial region, where, in the Aristotelian universe, transformations never occur. 
In order to extend his proof to cover the celestial region, Philoponus had to add 
that the sublunar and celestial regions are inextricably connected, so that the 
beginning of the one implies the beginning of the other. 59 The addition is no 
longer needed in Saadia's version of the proof, since the contention that existence 
could not have "traversed" infinite past time to reach the present moment applies 
equally to the sublunar and celestial regions. 60 

"Saadia, K. al-Amanat wa-l-l'tiqadat, ed. S. Landauer (Leiden, 1880), I, 1, p. 36; English 
translation with pagination of the Arabic indicated: Book of Beliefs and Opinions, trans. S. Rosenblatt 
(New Haven, 1948). 

56 Above, pp. 87-88. 

"Above, n. 11. Aristotle, Posterior Analytics I, 22, 83b, 6-7, states that it is impossible for an 
infinite to be "traversed in thought." 

58 Above, n. 8, and Aristotle, ibid. 59 Above. n. 10. 

'"There is at least one consideration that could make Philoponus' formulation stronger than 



Philoponus' Proofs of Creation 



97 



Saadia, at any rate, reveals indirectly that he did not recognize the difference 
between an infinite series and an infinite continuum. He takes up a possible 
objection to the proof, an objection which is, essentially, Zcno's first paradox, 61 
and which is attested to elsewhere in Arabic philosophic literature. 62 It runs: 
Since distance is infinitely divisible, would not the train of thought embodied in 
the proof render a person's moving from one place to another impossible, inas- 
much as the person would have to traverse an infinite number of parts? Saadia's 
response is that the objection mislcadingly adduces the traversing not of an "actually" 
existing infinite, but of an infinite existing solely in "imagination"; the proof, by 
contrast, rests on the fact that "existence does actually traverse [past] time and 
reach us," and an actual infinite cannot be traversed. 6 '' 

The objection faced here by Saadia views distance not as an infinite continuum, 
as Saadia's proof viewed time, but rather as an infinite scries of discrete parts. 
Since Saadia considers the objection to be pertinent, cither he is consciously 
drawing an analogy between an infinite scries of discrete parts and an infinite 
continuum, or else he did not detect any distinction between the two. In cither 
case the fact that Philoponus' proof addresses the former kind of infinite, whereas 
Saadia's addresses the latter, can occasion no hesitation about tracing Saadia's 
proof back to Philoponus; for Saadia, as now seen, does not recognize the dif- 
ference between the two kinds of infinite. 

Saadia's response to the possible objection to his proof incidentally reproduces 
a further detail from Philoponus: The principle that infinites cannot be traversed 
applies exclusively to actual infinites. 64 

(/'/') In a proof intended to establish that creation took place ex nihilo, Saadia 
reasons: If instead of accepting creation ex nihilo, "we were to suppose something 
coming from something, then the second thing in the hypothesis would resemble 
the first and would have to meet the condition that it can only come from a third 
thing. The third thing in the hypothesis would resemble the second, and would 
have to meet the condition that it can only come from a fourth thing. And this 
would go on to infinity. But since the infinite cannot be completed, ... we could 
not exist. Yet behold, we do exist! . . . whereas if the things preceding us were 
not finite [in number] they would not have been completed so that we might 
exist." 65 Saadia has, in a more abbreviated fashion, again offered the argument 



Saadia's, although Philoponus himself docs not state it: The assumption of an infinite scries of 
transformations would involve not only the traversal of an infinite; it would also involve an infinite 
regress of causes, which was rejected by Aristotle on independent grounds. Cf. below, p. 337. 

6l Cf., e.g., W. Ross's edition of Aristotle's Physics (Oxford, 1955), p. 72. 

62 Cf. below, p. 1 18. "Saadia, K. al-Amanat, I, 1, pp. 36-37. 

64 Cf. above, n. 7. On the question whether infinite past time would be an actual infinite, see 
below, p. 128. 

65 Saadia, K. al-Amanat, I, 2, p. 40. The same argument is alluded to in Saadia's Commentary 
on Sefer Yesira, published as Commentaire sur le Stfer Yesira, ed. M. Lambert (Paris, 1891), Arabic 
section, p. 3; French translation, p. 16. 



98 



Philoponus' Proofs of Creation 



that the present could not have been reached if, prior thereto, an infinite past had 
to be traversed. He has, in other words, again offered a version of Philoponus' 
first proof of creation from the impossibility of an infinite number. The argument, 
Saadia now claims, establishes creation ex nihilo, although the new version, like 
Saadia's previous version and like the original proof in Philoponus, docs not seem 
to preclude an eternal matter from which the world might have been formed. 66 
What is noteworthy about the new version is that it omits the details of Philo- 
ponus' proof which were reproduced in Saadia's previous version, while including 
two significant details that were absent there. The details appearing here but 
absent in the previous version are these: Saadia does not consider the time con- 
tinuum but rather, like Philoponus, the series of transformations leading up to 
what exists at the present moment. And he appends a statement to the effect that 
"if the things preceding us were not infinite, they would not have been completed 
so that we might exist." The statement appears to reflect Philoponus' contention 
that "if for the generation of a given thing there must first exist an infinite number 
of things that are generated from each other, then the given thing cannot be 
generated." 67 

Two items in the list of seven that I wish to examine in Saadia have been shown 
to derive from Philoponus's first proof from the impossibility of an infinite num- 
ber, namely, item (i), Saadia's fourth proof for creation, and item (ii), the argu- 
ment just discussed wherein Saadia claims to have demonstrated creation ex 
nihilo. 

(Hi) After presenting his proofs of creation ex nihilo Saadia takes up the unac- 
ceptable alternatives, among which is the theory that the heavens arc eternal and 
they bring the world into existence. In the course of refuting the eternity of the 
heavens he presents an argument entitled "from increase and diminution," which 
reads: "Every day elapsing from the time of the sphere is an increase over what 
has passed and a diminution from what is to come. But whatever admits increase 
and diminution is of finite power; and finiteness implies generation. Should any- 
one venture to assert that the elapsing of a day docs not increase what has passed 
or diminish what is to come, he would fly in the face of reality and experience ." <,x 
Saadia here has repeated the essential elements in Philoponus' second proof from 
the impossibility of an infinite number, the proof based by Philoponus on the 

66 The present version could establish creation ex nihilo only if one of Saadia's four proofs of 
creation — which I have designated as items (i), (v), (vi), and (vii) — had already proved that the very 
matter of the universe is created; but none of those proofs is represented by Saadia as doing so. In 
fact, the previous version of the proof under consideration — that is, item (i) — could make a stronger 
argument for creation ex nihilo than the present version, although Saadia represents the previous 
version as a proof merely of creation and the present version as a proof of creation ex nihilo. Should 
it be maintained that formless matter is properly described as existing in time, then it could be argued 
that since — in accordance with the previous version — past time must be finite, matter cannot have 
existed forever. 

67 Above, n. 9. 

68 Saadia, K. al-Amanat, I, 3 (eighth theory), p. 60. 



Philoponus' Proofs of Creation 



99 



principle that an infinite cannot be increased. As in the first item from Saadia 
which we examined, '''' the argument has been shifted into the realm of time. 
Saadia reasons that each passing day adds to the time already past; that the infinite 
cannot be increased; and that time must consequently be finite. Philoponus had 
made the same points regarding past "motions." 70 

There arc two elements in Saadia's proof which do not come from Philoponus. 
One is the statement that each elapsing day not merely adds to past time but also 
diminishes future time. The inexorable diminishing of future time contributes 
nothing to the proof for creation. But the consideration that what is subject to 
diminution cannot be infinite does appear in Arabic philosophy in a related con- 
text, in arguments for the impossibility of an infinite magnitude. 71 Saadia prob- 
ably intends by his statement to intimate an argument against the future eternity 
of the world. The thinking would be: Future time is steadily being diminished, 
the infinite cannot be diminished, therefore future time cannot be infinite. 

The other clement in Saadia's proof which docs not come from Philoponus is 
the conclusion that past time is "of finite power" ; we would have expected instead 
a conclusion to the effect that past time is of finite extent or of finite duration. 
Saadia surely has expressed himself poorly; for time cannot properly be described 
as possessing power, whether finite or infinite. Finite power was, as will be 
recalled, a central concept in a different proof of creation, the proof from the 
principle that a finite body can contain only finite power. And that proof, as will 
appear, was known to Saadia and employed by him. 72 We may conjecture that 
finite power intruded into the present proof as an echo from the other proof, 
where it originally belonged. Both proofs concern themselves with linitencss, 
although of different types, and Saadia — or someone earlier than he in the line 
of transmission — must have mechanically transferred the expression "finite power" 
from the proof where it is appropriate to a proof where it is not. 

(iv) Immediately after the argument just examined, Saadia continues his refu- 
tation of the thesis that the heavens arc eternal with an argument entitled "from 
the variation in movements." It reads: "Infinite power does not vary in itself. 
Since we observe that the movements of the heavens vary to the extent that they 
are thirty or 365 times one another, and still more, we realize that each |of the 
movementsj is finite. The explanation is as follows: The eastern movement of 
the highest sphere performs a revolution once in twenty-four hours, whereas the 
western movement of the fixed stars [i.e., the precession of the cquinoxes| pro- 
ceeds one degree each hundred years, at which rate the complete revolution will 
be performed in no less than 36,000 years or 13,140,000 days. . . . How can 
you say that a power whose movement varies so widely is not finite?" 7 ' 1 

""Above, pp. 95-96. 
"'Above, p. 88. 
71 Below, pp. 126-127. 
"Below, item (v). 

"Saadia, K. al-Amunat. I. 3 (eighth theory), pp. 60 - 6 1 . 



( 



100 Philoponus' Proofs of Creation 

i 

Partly explicitly and partly implicitly, Saadia has reproduced the key elements 
' in Philoponus' third proof from the impossibility of an infinite number, the proof 

( based on the impossibility of an infinite number's being multiplied. Like Philo- 

ponus, Saadia reasons that the heavenly bodies move at varying velocities; their 
' revolutions accumulate in varying multiples of each other; and therefore, "each" — 

l that is to say, the number of revolutions undergone by each of the heavenly 

bodies — must be "finite." 74 Saadia even remains faithful to the original proof in 
■ leaving the discussion within the realm of numbers of motions rather than trans- 

,/ ferring it — as he did in items (i) and (iii) — to the realm of time. There are small 

changes in detail: Saadia does not reproduce all the astronomical data given by 
' Philoponus. 75 In one instance, a substitution has been made; the figure "365" 

^ (the ratio of the revolutions of the diurnal sphere to the revolutions of the sun, 

and hence the number of days in a year), if not a scribal error, has been substituted 
( by Saadia or by his immediate source for the less familiar figure "360" (the ratio 

I of the revolutions of the moon to the revolutions of Saturn. 76 And an additional 

astronomical datum is evinced, with the unmistakable intent of magnifying the 

effect. The slow movement of the fixed stars, the phenomenon known today as 
I the precession of the equinoxes, had been calculated as one degree each hundred 

years. Accordingly, to the paradoxes of multiple infinites cited by Philoponus — 
' as reported by Simplicius — Saadia or his immediate source adds that the infinite 

^ revolutions of the daily sphere would, over an eternity, have to be no less than 

thirteen million times as numerous as the infinite revolutions of the fixed stars! 
\ As in the previous item, 77 Saadia speaks of finite power, the references to finite 

/ power appearing here not in the body of the argument but in an introductory and 

a closing statement. The introductory statement, which happens to have a close 
I parallel in Aristotle, 78 affirms that "infinite power does not vary in itself"; and 

^ the closing statement affirms that "a power whose movement varies" must be 

finite. These references to finite power are less incongruous than was the reference 



'•( to finite power in the previous argument. Movement may after all be described 

y as due to a finite or infinite power whereas time can hardly be described as 

! possessing either. Nevertheless, the introduction of finite power is again inap- 

( propriate. The present argument can be understood solely as reasoning from the 

\ f nature of number, from the impossibility of one infinite number's being a multiple 

I of another. The fact that the numbers of celestial revolutions accumulate in vary- 

j; ing multiples is pertinent when the argument is so understood; it supplies cogent 

| grounds for the conclusion that the number of revolutions of each heavenly body 

j is finite. That fact in no way exhibits, however, the finiteness of the power or 

( powers moving the spheres. The opening reference to finite power and the con- 



\( 

74 Cf. above, p. 89. 
; "Sec ibid. 

, 76 Ibid. "Above, pp. 98-99. 

'"Aristotle, De Caelo I, 7, 275b, 27-29. 



Philoponus' Proofs of Creation 



101 



eluding statement that "a power whose movement varies" is finite thus merely 
obscure the issue. The proper conclusion should be that the number of the rev- 
olutions of the heavens must be finite, and since the heavens cannot exist without 
revolving — as Saadia stales explicitly elsewhere''' — they cannot have existed 
forever. The concept of finite power, it may again be conjectured, has intruded 
into the present proof as an echo from a separate proof of creation, the proof 
from the finite power of the body of the universe. 

To recapitulate the discussion so far: The four items in Saadia which we have 
been examining are derived from Philoponus' three proofs for creation from the 
impossibility of an infinite number. The first of Philoponus' three proofs was the 
argument that no event could occur if, in order to reach it, an infinite series of 
transformations had to be traversed. The argument appears in Saadia in two 
versions, which complement each other in reproducing virtually all the details of 
Philoponus' proof. Philoponus' second proof was the argument that the number 
of past events must be finite since an infinite number cannot be added to; and his 
third proof was the argument that the number of the revolutions of the spheres 
must be finite since an infinite number cannot be multiplied. Those two proofs 
reappear in Saadia as part of a refutation of the eternity of the heavens. Saadia's 
versions reveal departures in detail from the original proofs, but the key elements 
of Philoponus' reasoning are preserved and Saadia even presents the proofs in 
the same sequence as Philoponus. 

(v) I turn to Saadia's first proof of creation. It is entitled "from finitude" and 
reads: "The heavens and earth clearly arc finite inasmuch as the earth is in the 
center and the revolution of the heavens goes around it. Hence their power is 
finite; for, as is well known, an infinite power cannot be present in a finite body. 
Since the power that maintains the heavens and earth ceases, they necessarily 
have a beginning and end." Saadia expands on the evidence for the finiteness of 
the heavens and earth 80 and concludes: "Since the bodies [of the heavens and 
earth] arc limited, their power is limited, reaching a certain limit where the power 
stops. They cannot continue after the destruction of that power nor exist before 
it docs." 81 

'"Saadia, A'. ol-Amunut. I, 1. p. 33. 

""Saadia explains thai the earth cannot be infinite, since il is circumscribed by the orbits of the 
heavenly bodies, and what can be circumscribed is not infinite. He explains that the heavens are not 
inlinitc, since the heavens revolve around the earth, and an infinite body could never complete a 
revolution. Thai argument is derived from Aristotle, lie CaeloX, 5, 271b. 26 ff. Subsequent to Saadia 
it can be found in the Kalam thinker Baghdad!, K. Usui al-Din (Istanbul, 1928). p. 66. Also cf. 
below, p. 257. Saadia finally explains that the universe cannot be infinite by virtue of being composed 
of an infinite number of heavens and earths. His grounds for ruling out a multiplicity of heavens and 
earths arc that each of the four elements can have no more than one natural place. That argument too 
is derived from Aristotle; cf. De Caelo 1, 8, and below, pp. 274-275. 

"'Saadia, K. al-Amanal, I, 1 , pp. 32-34. The proof is alluded to by Saadia in his Commentary 
to Sefer Yesira, pp 33 ff.; French translation, pp. 53 ff. 



102 



Philoponus' Proofs of Creation 



Here Saadia has reproduced Philoponus' proof of creation from the principle 
that a finite body can contain only finite power. The "well known" proposition 
cited by Saadia to the effect that "an infinite power cannot be present in a finite 
body" is exactly the Aristotelian principle upon which Philoponus' proof had 
been constructed. 82 Saadia establishes that the corporeal universe is in truth finite, 
and hence can, by the principle in question, contain no more than finite power. 
And he infers both the generation and the future destruction of the world on 
commonsense grounds: A finite power must eventually "cease" and "stop," and 
nothing can continue to exist after its power is exhausted. The inference reflects 
Philoponus' contention that what contains only finite power is subject to destruc- 
tion." 3 An indispensable step is, however, omitted by Saadia. Philoponus had 
explained that since the corporeal universe contains the possibility of not existing, 
it cannot be maintained in existence eternally even by an infinite power outside 
itself; for, over an infinite time, the possibility of not existing must, like every 
possibility, inevitably be realized. 84 Unless the step is included, the corporeal 
universe is shown by the argument merely to be incapable of maintaining itself 
eternally in existence through its own power. The universe is not shown to be 
absolutely incapable of existing eternally. Saadia's proof is, then, a simplified 
restatement of Philoponus' proof from the finite power of finite bodies, with an 
indispensable step missing. 85 

O0 Saadia's second proof for creation is entitled "from the joining of parts 
and the composition of segments." "Bodies," he contends, "consist of combined 
parts and composed members and thereby reveal . . . signs of generation and of 
the art of an artisan." As evidence, Saadia cites various types of composition 
discoverable in the corporeal universe, such as the combination of "earth, stones, 
and sand" to constitute the lower world, and the combination of the "several 
layers of spheres" and the "stars" to constitute the celestial region. He concludes: 
"After establishing joining, conjunction, and composition, which arc generated 
things (hawaditli) in the body of the heavens and the rest [of the universe], 1 was 
convinced . . . that the heavens and everything they contain arc generated." 86 

What is significant for us in this proof is its location. Saadia presents it imme- 
diately after his proof from the finite power of finite bodies. Philoponus, as will 
be recalled, had supported his own proof from the finite power of finite bodies 
with a number of auxiliary arguments, intended primarily to show that the cor- 
poreal universe can indeed contain only finite power. The first, and also the third, 
of the auxiliary arguments reasoned from the composite nature of the corporeal 



82 Above, p. 89. 
"Above, p. 90. 
84 lbid. 

85 H. Wolfson, "Kalam Arguments for Creation," Saadia Anniversary Volume (New York, 1943), 
p. 203, offers an interpretation of Saadia, the effect of which is to discover in Saadia the technicalities 
of Philoponus' proof. 

86 Saadia, K. al-Amanal, I, I, p. 34. Wolfson, "Kalam Arguments for Creation." p. 205. n. 34, 
notes that the term pans at the beginning of the argument might mean atoms. 



Philoponus' Proofs of Creation 



103 



universe. Thus Philoponus' proof from the finite power of finite bodies was 
followed immediately by an argument from the fact that all bodies in the universe 
are composite; and Saadia's proof from the finite power of finite bodies is now 
also followed immediately by an argument from the fact that all bodies in the 
universe arc composite. 

In content, Saadia's argument from composition differs from Philoponus'. The 
evidence cited by Saadia for the composite nature of the corporeal universe is 
not the evidence cited by Philoponus, 88 and there is no hint in Saadia of a 
connection between composition and finiteness of power. Yet in whatever light 
Saadia's proof might be regarded, 89 Saadia undoubtedly has omitted something 
from his source or left part of his thinking unexpressed, for he fails to explain 
how the premises lead to the conclusion that the world was created. He merely 
takes for granted that if things are composite they must be generated. The only 
intimation of grounds for a conclusion is given by him in the statement: "Con- 
junction and composition . . . are generated things in the body of the heav- 
ens. . . ." The thought behind the statement plays an important role in another 
proof for creation offered by Saadia, the proof to be examined next. The appear- 
ance of the same thought here is perhaps an additional instance of the intrusion 
of echoes of one proof into another; alternatively, it is a conscious attempt to 
assimilate the present proof to the other in order to justify the conclusion. 9 " But 
however that thought came to be introduced, it cannot represent the original point 
of the argument, for then there would be no reason for distinguishing between 
the present proof from "composition" and the succeeding one "from accidents." 

O'O Saadia's third proof for creation is, as just mentioned, entitled "from 
accidents." It begins: "Bodies cannot avoid accidents which occur in each of 
them, either from itself or from outside itself." 91 Saadia examines animal life and 
the earth as a whole, discovers that they arc indeed "not free" of change, which 

"'Above, pp. 92-93. 

88 Abovc, p. 92. Saadia's failure to mention the composition of matter and form might be explained 
by the fact that the proof came to him after it had been reformulated by Kalam thinkers who rejected 
the Aristotelian matter and form. 

"'Several arguments for creation from composition were put forward in the Middle Ages; cf. 
below, pp. 146 If. 

'"' That is suggested by M. Ventura. In Philosophic tie Siuuiio (loon (Paris. 1934). p. ')'). 

'"The phrase "cither from itself or from outside itself" can be deciphered with the aid of for- 
mulations of the proof in other Arabic writers. As a premise, it was common procedure to demonstrate 
the Kalam doctrine of accidents, the doctrine that none of the characteristics of a physical object flow 
from, or are dependent upon, an inner essence, but rather all are added to the identical inert atoms 
that serve as the material base of each object. To establish the existence of accidents it was shown 
that the characteristics of things cannot be due to the atom or body itself, and therefore must be due 
to a factor distinct from the atom or body, that is. to an "accident"; cf. below, p. 181. By staling 
that the accidents he is speaking of do belong to a body "either from itself or from outside itself." 
Saadia explicitly rejects the proposition that no characteristic of a thing can flow "from itself" and 
he implicitly rejects the Kalam doctrine of accidents. He would, we may suppose, have been more 
comfortable with the Aristotelian theory of form. But he apparently did not have the philosophic 
knowledge to allow him its use; cf. Ventura. La Philosophic tie Saailia. pp. 102-103. 



I. 

104 Philoponus' Proofs of Creation 

i 

f is a "generated thing," and concludes that "what cannot avoid what is generated 

is known to be of the same character." He hereupon extends his investigation to 
! the heavens, discovers that they too are not free of motion and other accidents, 

/ and concludes again: "When 1 discovered that generated things embrace these 

bodies, and the latter do not precede the former [in time], I was convinced that 
( whatever does not precede what is generated is of the same character by virtue 

of its falling under the same class (haddihi)." 92 
Saadia's proof addresses accidents and bodies, while a parallel version of the 
( proof known from Islamic writers addresses not accidents and bodies but acci- 

^ dents and atoms. 93 The dichotomy of accident and body, or of accident and atom, 

is a Kalam analogue of the familiar Aristotelian dichotomy of form and matter: 
( The Aristotelian concept of a form, which carries the essential nature of a given 

^ object, was rejected by the Kalam, and whatever trace it left was absorbed into 

the Kalam concept of accident. 94 The specific phenomenon that Saadia's proof 
I considers is the continuous and unavoidable presence of generated accidents in 

each body in the universe; the phenomenon that the parallel version considers is 
1 the continuous presence of generated accidents in each atom. That phenomenon, 

( translated back into the Aristotelian framework, would be the continuous presence 

of generated forms in matter. As will be recalled, the second of the auxiliary 
' arguments with which Philoponus supported his proof from the finite power of 

( finite bodies focused precisely on the succession of forms across matter. 95 Sa- 

adia's proof thus corresponds, in a most general way, to Philoponus' second 
* auxiliary argument: Philoponus dealt with the continuous succession of forms 

/ over matter, while Saadia deals with the continuous presence of changing gen- 

erated accidents in bodies. The similarity goes no further. Saadia does not repeat 
' any details from Philoponus; and his argument rests on a problematical rule, not 

/ reflecting anything in Philoponus, namely, the rule that "what cannot avoid what 

is generated is known to be of the same character." 96 
( The situation we have discovered is this: Philoponus had offered a proof from 

/ the finite power of finite bodies followed directly by a supporting argument from 

C ,2 Saadia, K. al-Amanat, I, 1, p. 35. The proof also appears in Saadia's Commentary to Sefcr 

Yesira, pp. 33 ff.; French translation, pp. 53 ff. It is alluded to by Saadia in his refutation of Hiwi; 

^ cf. 1. Davidson. Saadia's Polemic Against lliwi Al-Ualkhi (New York, 1915), p. 75, §65. 

The last nine words in the proof are understood by Ventura, Rosenblatt, and Altmann as 
meaning: "by virtue of its [i.e., that which is generated] entering its definition [i.e., the definition 

f of that which does not precede.]." This would seem to require the following, rather forced, interpre- 

tation of the text: Since the concept accident is used by the Kalam in defining body, and since all 
accidents are generated, it follows that "what is generated" enters the definition of body, and body 
is also generated. Definitions of body in terms of accident are in fact to be found in Ash'ari, Maqalat 

( al-Islamtyin, ed. H. Ritter (Istanbul, 1929-1930), pp. 301 ff., and the first of the definitions given 

there would be particularly appropriate for our context. 

' 93 Cf. below, pp. 134 ff. 

94 Cf. Maimonides, Guide, I, 73(8); R. Frank, The Metaphysics of Created Being According to 

■ Abu l-Hudhayl (Istanbul, 1966), p. 42. 

"Above, p. 92. ''Regarding that rule, see below, p. 143. 



Philoponus' Proofs of Creation 



105 



the composition of all bodies in the universe and thereupon by a supporting 
argument from the unceasing succession of impermanent forms over matter. 
Saadia, for his part, offers a proof from the finite power of finite bodies followed 
directly by a proof from the composition of all bodies in the universe and there- 
upon by a proof from the continuous presence of impermanent accidents in bod- 
ies. Saadia's proof from the finite power of finite bodies is unmistakably derived 
from Philoponus. The similarity of Saadia's second and third proofs to the cor- 
responding auxiliary arguments in Philoponus is limited. Saadia's proofs consider 
the general phenomena considered by Philoponus' proofs, and likewise appear 
directly after the proof from finite power, but they differ from Philoponus' proofs 
in all particulars. 

In each of the three proofs, Saadia's reasoning is incomplete. His restatement 
of the first proof, the proof from the finite power of finite bodies, omits a step 
without which the conclusion cannot cogently be drawn. 97 As he formulates his 
second proof, the inference of creation from composition is not explained. And 
his third proof, the proof "from accidents," rests on the unproved rule affirming 
that "what cannot avoid what is generated is known to be of the same character." 
The proof from accidents must already have been current in Arabic by Saadia's 
time, since two of his contemporaries, Ash'ari 98 and Alfarabi, 99 expected their 
readers to be acquainted with it. Ash'ari significantly attributes the proof to the 
"philosophers" and "those who follow them." 100 

I would suggest that Philoponus' proof from finite power may have circulated 
in Arabic together with at least two 101 auxiliary arguments as a group; and that 
group of proofs underlies Saadia's first, second, and third proofs of creation, the 
items I have designated as (v), (vi), and (vii). If the suggestion is correct, far- 
reaching changes must have occurred somewhere along the line of transmission: 
The connection between the auxiliary arguments and the basic proof was forgot- 
ten; essential parts of the reasoning were omitted; and in the case of the third 
proof, the original concepts were translated into their Kalam analogues and pro- 
vided with a completely new line of argumentation. If the suggestion is not 
correct, pure chance is the only explanation for what has been shown, namely, 
that the two proofs following the proof from finite power in Saadia exhibit a 
correspondence to the two auxiliary arguments following the proof from finite 
power in Philoponus. 

'"Above, p. 102. 
"•"Sec n. 100. 

w See below, pp. 134-135. 

'""Ash'ari, Risala Ha ahl al-Thaghr. Publications of the Theological I-acully. Istanbul, 11/8 (1928). 89; 
cf. R. Frank, "Al-Ash' art's Conception of the Nature and Role of Speculative Reasoning in Theology." 
Proceedings of the Villi Congress of Arabic and Islamic Studies (Stockholm, 1972), p. 139. Ash'ari, 
more specifically, attributes the argument to "the philosophers, and those who follow them from 
among the qadariya. the innovators [or: heretics], and the deviators from the prophel(s)." He probably 
has the Mu'lazilites in mind. 

""That is to say. the first two or first three auxiliary arguments. 



106 



Philoponus' Proofs of Creation 



Resume™ 2 

Philoponus' three proofs for creation from the impossibility of an infinite num- 
ber were known to and used by Saadia (items (i)-(iv)|. 103 Philoponus had pre- 
sented the three proofs as a set, but whether Saadia received them as a set is not 
certain. Philoponus' proof from the finite power of finite bodies was also known 
to and used by Saadia [item (v)]. 104 After his proof from finite power, Saadia 
offered two other proofs [items (vi) and (vii)| which in their general subject 
matter and order of presentation correspond to the first two auxiliary arguments 
whereby Philoponus had supported his own proof from finite power. 105 The 
resemblance, I have submitted, is not fortuitous. 

An interrelation is likely as well between the transmission of Philoponus' proofs 
from the impossibility of an infinite number and the transmission of his proof 
from finite power; for in Saadia, two proofs from the former group betray echoes 
of the latter. 106 Quite possibly, therefore, a list of six — or more — proofs for 
creation, all of them concerned with finiteness of one sort or another, may have 
circulated and been available to Saadia. The list would have consisted of Philo- 
ponus' three proofs from the impossibility of an infinite number, his proof from 
finite power, and at least two auxiliary arguments supporting the proof from finite 
power. 



3. Kindi and Philoponus 

Several generations before Saadia, Kindi formulated four arguments that also 
exhibit unmistakable traces of Philoponus. For the purpose of exposition, 1 shall 
disregard the order in which the arguments appear in Kindi. The following table 
of corresponding items may again be helpful. 



Kindi 



Philoponus 



(i) Argument for the finiteness 
of past time 

(ii) Argument for the finiteness 
of bodies 

(iii) Another argument for the 
finiteness of past time 

(iv) Argument from 
composition 



(a.i) The infinite is not 

traversable 
(a.ii) The infinite cannot be 

added to 
(a.ii) The infinite cannot be 

added to 

(b.ii, b.iv) Auxiliary arguments in 
support of proof from 
finite power 



; Sce tabic, above, p. 105. 
'Above, pp. 95-99. 
'Above, p. 101. 
'Above, pp. 102-104. 
'Above, pp. 99, 100. 



Philoponus' Proofs of Creation 



107 



(i) The first item in Kindi to be examined is a proof of the finiteness of time, 
which reads: "|If past time were infinite], then before every segment of time 
there would be another segment, ad infinitum. But in that case no given time 
could ever be reached. For [ the duration] from the infinite past to the given time 
would be equal to the duration ascending back from the given time ... to 
infinity. . . . If something is infinite, its interval cannot be traversed." 107 After 
proving that past time cannot be infinite, Kindi explains that time cannot become 
"actually infinite" in the future either; for no matter what "definite time" might 
be added to the already accumulated finite past time, the total must remain finite. 108 

Kindi is plainly offering another restatement of Philoponus' first proof for 
creation from the impossibility of an infinite number. Like Philoponus, he lays 
down the rule that an infinite cannot be traversed; he draws an analogy between 
going back up through the infinite past and coming down from the infinite past 
to the present; and he concludes that the present could never have been reached 
if an infinite past were to precede it. 109 Kindi departs from Philoponus in trans- 
ferring the discussion from the realm of past transformations — the realm consid- 
ered by Philoponus 1 10 — to the realm of past time, something that Saadia, as will 
be recalled, also did in his restatement of Philoponus' proof. 1 1 1 Kindi's departure 
from Philoponus is not, however, identical with Saadia's, since Kindi, unlike 
Saadia, does not treat past time as a continuum but instead as a series of intervals. 
The conception of time as a series of intervals is quite atypical and would appear 
to serve a single purpose: It allows Kindi to argue specifically against an infinite 
past scries, as Philoponus had done. Another similarity between Kindi's version 
and Saadia's, in addition to transferring the discussion to the realm of time, is 
that both versions supplement Philoponus' proof for the finiteness of past time 
with a proof, or at least a statement," 2 of the finiteness of future time. The 
supplement indicates that Kindi and Saadia borrowed from a common line of 
transmission in which the finiteness of future time — a theme appearing in Phil- 
oponus' arguments for creation from finite power — had attached itself to Philo- 
ponus' proof for the finiteness of past time. Also to be observed arc clear ter- 
minological similarities between Kindi and Saadia. Both use the same distinctive 
terms for ascending in time (K: mutasa' idan\ ni S: su'ud UA ) and traversing 
time (K: Iii tuqta; S: lam yaqta'hu " 5 ). 

" ,7 Kindi. Rasa'il. eel. M. Abu Rida (Cairo. 1950). 1. p. 121. Parallel versions appear, ibid., pp. 
197; 205-206. English translation, with pagination of the Arabic indicated: Al-Kindis Metaphysics. 
trans. A. Wry (Albany, 1974). 

""Kindi, Rasa'il. 1. p. 122. See below, n. 121; Aristotle, Physics III, 6, 206a. 25-29. 

'""Above, pp. 87-88. 

""Above, p. 87. 

'"Above, pp. 95-96. " 2 Above. p. 96. 

" 3 I have given the reading of a parallel passage, Rasa'il, p. 197. The editor of Kindi's Rasa'il 
did not understand the point of the argument and made an incorrect emendation in the present passage, 
p. 121, 1, 10, and note. 

""Saadia, K. al-Amanal, I, 1, p. 36, and above, p. 96. '"ibid. 



108 



Philoponus' Proofs of Creation 



Kindi, then, knew and employed Philoponus' first proof lor creation from the 
impossibility of an infinite number and very likely received it through the same 
line of transmission from which Saadia was subsequently to borrow. 1,6 

(;7) A second argument in Kindi begins by establishing the linitcness of the 
body of the universe, whereupon it proves the linitcness of time as a corollary. 
In establishing the finiteness of the body of the universe, Kindi writes: "Supposing 
that an infinite body does exist, a finite [portion of it] may be assumed to be 
removed . . . [and subsequently] restored." Now, Kindi continues, the remainder 
left after a finite portion was removed could be neither finite nor infinite. The 
remainder could not be finite; for when the subtracted portion is restored, the 
total, as the sum of two finite magnitudes, would also be finite, whereas the 
initial supposition was that the total is infinite. The remainder left after a finite 
portion is removed could just as surely not be infinite. The remainder could not 
be infinite and yet, on the one hand, fail to be increased when the subtracted 
portion is restored; for every magnitude undoubtedly becomes greater when another 
magnitude is added to it. Nor could the remainder be infinite and yet, on the 
other hand, be increased through the restoration of the finite portion; for "there 
cannot exist two infinite magnitudes of the same type" 7 one of which is greater 
than the other." Since the remainder left after removing a finite portion could be 
neither finite nor infinite, an infinite body, Kindi concludes, cannot possibly 
exist." 8 

The argumentation turns on the proposition that "there cannot exist two infinite 
magnitudes of the same type one of which is greater than the other." That prop- 
osition is equivalent to the principle lying at the heart of Philoponus' second 
proof of creation from the impossibility of an infinite number, the proof in which 
Philoponus had reasoned that the number of past motions leading up to the present 
cannot be infinite, since the number of such motions is subject to increase whereas 
an infinite cannot conceivably be increased; 1 19 to affirm, as Kindi does, that one 
infinite cannot be greater than another is equivalent to affirming that the infinite 
cannot be increased. Kindi 's procedure of mentally removing and restoring a 
portion of the supposed infinite is designed to bring the principle in question to 
bear upon an infinite body. Philoponus needed no such device because he was 
considering the number of motions that have taken place in the universe, and the 
number of motions is, manifestly, ever growing. Kindi, though, is considering a 
body. He cannot, for his part, flatly assert that every body is subject to increase, 

1 ""Saadia docs not draw his proof from Kindi, for he has details of Philoponus' proof which Kindi 
does not have. Neither Kindi nor Saadia drew their proofs from the version published by Pines, "An 
Arabic Summary of a Lost Work of Philoponus," pp. 330-336 (above, n. 52). For they include key 
points — the impossibility of traversing an infinite; and the comparison between going back up through 
the infinite and coming down to the present — which that version omits. 

" 7 That is to say, two infinite lines, two infinite surfaces, or two infinite bodies. 

""Kindi, flora' (7, I, pp. 1 15-1 16, taken together with pp. 189-191 ; also cf. pp. 195.202-203. 

'"Above, p. 88. 



Philoponus' Proofs of Creation 



109 



anil (hereupon argue that an infinite body would be subject to increase, which is 
impossible; for the obvious response would be that infinite bodies are not after 
all subject to increase inasmuch as nothing exists outside them which might be 
added to them. The procedure of mentally removing and subsequently restoring 
a portion serves to reveal that the supposition of an infinite body docs in truth 
involve an infinite's being increased: The remainder left after subtracting a finite 
portion could not be finite — since the total would then be finite — but would have 
to be infinite; and when the subtracted portion is restored, an infinite body would 
have been increased. Once the supposition of an infinite body is known to involve 
an infinite's being increased, Kindi can reason, as Philoponus did in connecfion 
with past events, that an infinite cannot conceivably be increased and conse- 
quently an infinite body is impossible. 

Kindi 's proof of the impossibility of an infinite body is plainly an adaptation 
of the argument whereby Philoponus had proved the impossibility of an infinite 
number of past motions. 

In the context where the argument appears, Kindi is not interested in the 
finiteness of bodies for its own sake. He is constructing a proof of creation and, 
within the economy of the overall proof, the finiteness of bodies is intended to 
prepare (he ground for its more important corollary, the finiteness of time. Kindi 
has two ways of deriving the corollary. 

In several passages he contends that inasmuch as every body is finite, and 
inasmuch as time is an accident of body, time too must be finite. 120 His reasoning 
is highly questionable, the sense in which a body is finite being entirely different 
from that in which time would be finite; time extends in a different dimension 
from the three dimensions in which bodies extend. It is difficult, therefore, to 
sec how the finiteness of the three dimensions of body can imply the finiteness 
of the temporal dimension. 

(Hi) Side by side with the foregoing contention, Kindi advances the more 
apposite contention that the "method" establishing the finiteness of bodies may 
be employed comprehensively to "show that nothing quantitative can be actually 
infinite"; 121 and inasmuch as "time is quantitative, ... an actual infinite time is 
not possible." 122 In other words, the supposed infinite past time would be subject 
to increase, whereas an infinite cannot conceivably be increased; and hence past 
time cannot be infinite. Here we have Philoponus' second proof for creation from 
the impossibility of an infinite number transferred — as in Saadia 123 — from the 
realm of past motions to the realm of past time. 

Both Philoponus' first and second proofs of creation fom the impossibility of 
an infinite number thus reappear in Kindi, the discussion in both instances being 

,20 Kindi, Rasa il, I, p. 116; cf. ibid., pp. 196, 203. 

l2l The potentially infinite, in contrast to the actually infinite, consists, as Kindi explains, in the 
possibility of continually adding to a given quantity. Cf. Rasa'il, I, p. 1 16; Aristotle, Physics 111, 6. 
'"Kindi, Rasa'il, I, p. 116. 
'"Above, pp. 98-99. 



110 



Philoponus' Proofs of Creation 



transferred to the realm of past time: Philoponus had contended that transfor- 
mations presently taking place could never have been reached if, prior thereto, 
an infinite scries of transformations had to be traversed; and again that an infinite 
number of past motions is impossible since the number of past motions is con- 
stantly increased whereas the infinite cannot be increased. Kindi contends litem 
(i)] that the present segment of time could not have been reached if, prior thereto, 
an infinite scries of segments had to be traversed; 124 and again [item (iii)] that 
an infinite past time is not possible, since past time — like other quantities — is 
subject to increase whereas the infinite cannot conceivably be increased. 

One might have expected Kindi to regard his proofs of the finiteness of past 
time as self-contained proofs for the creation of the world. In Philoponus, a proof 
of the impossibility of an infinite number of past transformations or past motions 
sufficed as a proof of the creation of the world. And in Saadia a proof of the 
impossibility of infinite past time also sufficed as a proof of creation, since it 
went without saying for him that if past time is finite, the world cannot have 
existed forever. Yet Kindi is not satisfied that his proof of the finiteness of past 
time is tantamount, by itself, to a complete proof of creation. He feels called 
upon to show explicitly that the finiteness of past time docs indeed imply a 
beginning of the world, or, to use his words, that it implies the finiteness of the 
body of the universe in respect to "existence" (iimiya ). 125 Kindi's procedure for 
showing that the finiteness of past time does imply creation consists in demon- 
strating the coexistence of the physical universe and time. Again, he has various 
ways of accomplishing his task. 

In a certain passage he reasons that if, on the one hand, the body of the universe 
is not eternal, its very generation was a motion, and therefore the body of the 
universe and its motion are coexistent; and if, on the other hand, the universe is 
eternal, it must always have been in motion, for were it ever completely at rest, 
it could never have started to move. 12(1 Whether eternal or not, the body of the 
universe is, then, coexistent with motion. 127 Time is likewise coexistent with 
motion, since by definition time is the "number of motion." 128 "The body |of 
the universe], motion, and time" are, consequently, coexistent with each other 
and "do not ever precede one another." 129 Kindi's previous arguments to the 
effect that "all time is finite" and his present argument to the effect that "body 
docs not precede time" lead to the desired conclusion that "the body of the 
universe cannot be infinite in its existence (inniya)," but must have a beginning. 



l24 Above, p. 107. 
'"Below, n. 130. 

l26 The contention that if the universe were eternally at rest it could never begin to move is 
Aristotelian; cf. Physics VIII, 1, 251a, 20 ff. The contention is to be found in the Kalam; cf., e.g., 
Fakhr al Din al Razi, K. al-Arba'in (Hyderabad, 1934), p. 16. 

l27 Kindi,/?ajd'i7. 1, pp. 118-119. 

,28 Ibid., p. 1 17; cf. Aristotle, Physics V, 1 1 , 220a, 24-25. 

129 Kindi, Rasa' it, I, p. 119. 

I30 lbid., p. 120; cf. pp. 197-198, 204-206. 



Philoponus' Proofs of Creation 



111 



The reasoning has an Aristotelian cast, 131 and is straightforward enough. Else- 
where, however, Kindi establishes the coexistence of body, motion, and time in 
a convoluted and much more problematic fashion. 132 Embodied within the con- 
volutions, another of Philoponus' proofs can, 1 think, be detected. 

(iv) Kindi writes: "Composition and combination are change, for this [last] 133 
is a joining and ordering of things. Body is a long-wide-decp, or three dimen- 
sional substance, and hence is composed of substance, its genus, and long-wide- 
decp, its specific difference. It [also] is what is composed of matter and form. 
Now composition is the change from the state that is noncomposition. Compo- 
sition is motion. . . . Therefore the body [of the universe] and motion do not 
precede one another. . . . Ami accordingly the body |of the univcrsc| does not 
precede duration numbered by motion [i.e., timc|. Therefore the body [of the 
universe], motion, and time do not precede one another in existence (/nniya), 
but are simultaneous in existence. Since time was |already proved to be| finite in 
actuality, the existence of the body [of the universe] also is necessarily finite in 
actuality, inasmuch as composition and combination are a certain change. If, on 
the contrary, composition and combination were not change, the conclusion would 
not be necessary." 1 34 

Apparently Kindi is reasoning that the universe is a body; every body is com- 
posite, since it consists of matter together with form, or substance together with 
tridimcnsionality; composition is a change from the state of noncomposition; 
change is a type of motion; hence the body of the universe cannot have existed 
before motion but is coexistent therewith; motion and time are likewise coexistent; 
hence the universe cannot have existed before time; but time is finite; hence the 
body of the universe can have existed for only a finite duration. 

If such is indeed Kindi's reasoning, it contains a weak link. Kindi lays down 
that bodies and the universe as a whole are composite (murakkab) and infers that 
the body of the universe has undergone change and motion, since composition 
(larkib) is a change. The inference would appear to be accomplished through a 
blatant equivocation. In Arabic as in English, composition can mean both a state 
and a process; and Kindi appears to be gliding from composition in the universe, 
in the sense that the universe exhibits the state of being composite, to composition 
in the universe, in the sense that the universe has undergone the process of 
becoming composite. Only thus is he able to infer that the universe has undergone 
change. To rescue Kindi from the equivocation, we may perhaps understand him 
to have assumed a hidden premise, the proposition that everything in a state of 
being composite must have undergone a process rendering it so. The hidden 
premise does not, however, greatly improve the argument. For the premise is far 

'"Cf. Aristotle, Physics IV, 12. 

'"Rasa'il. pp. 196-197. offers yet another argument for the coexistence of body and time; the 
argument is less well developed and is diflicult to follow. 
'"I have taken "change" as the antecedent of "this." 
" 4 Kindi, Rasail. I, p. 120; cf. ibid., p. 204. 



( 

( 

^ 112 Philoponus" Proofs of Creation 

( 

from self-evident and, if employed by Kindi, certainly should have been made 

explicit and defended. 135 
/ Moreover, quite apart from the equivocation or hidden premise, Kindi 's rea- 

f soning is curiously circular. He adduces the fact that the universe has undergone 

a process of composition, as a single link in an elaborate chain of argumentation 
i leading to the creation of the world. But once the body of the universe, being 

composed of "matter and form," is acknowledged to have undergone a process 
' of composition — from, as Kindi puts it, "the state ... [of] noncomposition" 136 — 

( the body of the universe is immediately acknowledged to have come into exis- 

tence. Kindi is, in effect, contending that given the premise that the body of the 

universe came into existence and several other premises, the coexistence of the 
/ universe and time ensues; and given the coexistence of the universe and time, 

his earlier proof of the finiteness of past time will imply that the body of the 
' universe — came into existence. The premise that the body of the universe came 

I into existence allows him to conclude, after a number of intermediate steps, that 

the body of the universe did in truth come into existence. 
' The most plausible explanation of what has occurred is this: Kindi must have 

I known of an argument proceeding from composition to creation but did not 

understand or remember exactly how it ran. 137 As a result, he incorporates the 
' argument from composition into the proof of creation from the finiteness of time, 

^ employing it merely as a device to help show that the finiteness of past time does 

imply creation. 

( The evidence for composition adduced by Kindi indicates the probable source 

/ of his argument for composition. Body, Kindi finds, is a "three dimensional 

substance, hence [it] is composed of substance, its genus, and long-wide-deep, 
I its specific difference"; furthermore, body is "composed of matter and form." 138 

( Philoponus, as will be recalled, had supported his proof for creation from the 

finite power of the universe with several auxiliary arguments, the first and third 
f of which took their departure from the presence of composition in the universe. 

, The evidence for composition adduced by Philoponus in the first of the auxiliary 

arguments was the circumstance that "the heavens are [composed] of matter and 
( form." 139 The evidence adduced by him in the third auxiliary argument was, 

f again, that the heavens are "composed of substratum . . . and form." And Phil- 

oponus added thereto the observation that even should the distinction of matter 

i 

( '"The premise might, conceivably, be alluded lo in the first and last sentences of the passage. 

See also below, p. 148. 
< '"Above, n. 134. 

l37 Kindi, it should be mentioned, is reported elsewhere to have deduced nonetcrnity directly from 
' composition. According to the report of Ibn 'Adi, he refuted the Christian trinity by contending: 

! "Everything composite is caused, and everything caused is nonetcrnal." Cf. the text published in 

Revue de L'Orienl Chretien, XXII (1920), 4, line 15; French translation in A. Pcricr, Pclits Traites 
( Apolngetiques de Yahya Ben 'Adi (Paris, 1920), p. 119. 

''"Above, n. 134. "'Above, p. 92. 

( 



Philoponus' Proofs of Creation 



113 



and form in the heavens be disputed, the presence in the heavens of "extension 
in three dimensions" — which is a mode of composition— must still be acknowl- 
edged. 140 The aspects of composition adduced by Kindi turn out to be precisely 
those that had been adduced by Philoponus in his auxiliary arguments from 
composition. It seems probable, therefore, that Kindi knew one or both of Phil- 
oponus' arguments for creation from composition in the universe, but, at least at 
the time he was writing, did not recall them exactly. He adduces and utilizes the 
evidence of composition in the universe not in order to formulate an independent 
proof for creation nor, like Philoponus, to formulate an auxiliary argument in 
support of the proof of creation from the finite power of the universe. Instead, 
he utilizes the evidence of composition for the purpose of showing that the 
finiteness of time does imply the creation of the world. It is worth mentioning 
also that although Kindi employs his argument from composition differently, a 
certain similarity to the overall configuration in Philoponus can be discerned: 
Kindi too employs the argument from composition in support of another proof 
for creation, rather than independently. 

To recapitulate: Kindi (i) establishes the finiteness of past time through a 
restatement of Philoponus' first proof of creation from the impossibility of an 
infinite number. The argument is that an infinite cannot conceivably be traversed, 
and consequently the present moment could not, on the assumption of eternity, 
ever have been reached. Kindi (ii) establishes the finiteness of all bodies through 
an adaptation of Philoponus' second proof of creation from the impossibility of 
an infinite number; the argument here rests on the principle that an infinite cannot 
conceivably be increased. Once Kindi proves the finiteness of all bodies, he 
derives therefrom the corollary that past time must be finite. His reasoning is that 
since time is an accident of body, and the body of the universe is finite, time too 
must be finite. And, he adds more appositely, (iii) the principle that an infinite 
cannot be increased may be applied to time as well as body, thereby establishing 
that time too is finite. Kindi has thus recast Philoponus' first and second proofs 
of creation from the impossibility of an infinite number as proofs for the finiteness 
of past time. For Kindi, a proof of the finiteness of time docs not yet, by itself, 
constitute a proof of creation, and he feels called upon to show that the finiteness 
of time does imply the finiteness of the "existence" of the body of the universe. 
One device he employs to that end is (iv) an argument from composition. Exam- 
ination of his argument from composition discloses, however, that it cannot 
originally have been intended for the end to which Kindi employs it; and, further, 
that it cites the very evidence for composition which had been cited by Philoponus 
in his auxiliary arguments from composition. Philoponus is therefore again indi- 
cated as Kindi 's likely ultimate source.' 41 



""A connection between Kindi and Philoponus was also observed by R. Walzer, Greek into 
Arabic (Oxford. 1962), pp. 190-196. 



114 



Philoponus' Proofs of Creation 



Earlier we saw that Saadia offers adaptations of at least four of Philoponus' 
proofs for creation, including Philoponus' first and second proofs from the impos- 
sibility of an infinite number. 142 Those proofs have now been found in Kindi as 
well, transferred, as in Saadia, to the realm of time. In addition, Saadia, like 
Kindi, offers an argument for creation from composition in the universe, and 
both writers present their arguments from composition in association with other 
proofs for creation deriving from Philoponus. 143 The reasoning of Saadia's argu- 
ment from composition differs from the reasoning of Kindi's; but in both instances 
the reasoning exhibits such obvious gaps that the arguments must be incomplete 
restatements of a once known, half-forgotten earlier proof. 144 Significantly, Sa- 
adia and Kindi use the same three terms — composition (tarkib), combination (K: 
i'tilaf; S: ajza' mu'allafa), and joining (jam') — in formulating their argu- 
ments. 145 With regard to Saadia, I suggested that a list or collection of proofs 
for creation deriving from Philoponus circulated and was utilized by him. I would 
here suggest that Kindi knew a collection of proofs for creation deriving from 
Philoponus which resembled that known to and utilized by Saadia. The collection 
available to Kindi included at least Philoponus' first and second proofs from the 
impossibility of an infinite number and one or both of his auxiliary arguments 
from composition. Either the form in which the arguments from composition 
reached Kindi did not permit him to grasp the original point, or else Kindi forgot 
the point when he sat down to write. He was consequently left — like Saadia — 
to employ the argument as best he could. 

Two small details in Kindi's discussion of creation may cast added light on his 
source. The first detail is this: Each time Kindi takes up the issue of creation, he 
opens with the proof of the finiteness of bodies [item (ii)J, whereupon he proceeds 
to a proof of the finiteness of past time. 146 The proof of the finiteness of bodies 
contributes only superficially to the proof of the finiteness of past time. The 
finiteness of bodies prepares the ground for the highly questionable inference that 
since time is an accident of body, time too must be finite. 147 And the proof of 
the finiteness of bodies allows Kindi to observe that the same argumentation — 
an infinite cannot be increased, hence what is subject to increase cannot be 
infinite — will apply to everything quantitative, so that time too can be proved to 
be finite. 148 A much simpler procedure would plainly have been to apply the 
argumentation immediately to time without reference to the finiteness of bodies. 
But the finiteness of bodies did serve as an essential premise in a different proof 
of creation, Philoponus' proof from the finite power of finite bodies. 149 In Phil- 
oponus' own formulation of that proof the finiteness of the body of the universe 



Above, pp. 95-102. 
'Cf. above, p. 102. 

'Cf. above, p. 103. 145 Above, nn. 86 and 134. 

s Kindi, Rasa'il, I, pp. 115-116; 195-197; 203-204. 
'Above, p. 109. 

'ibid. ""Above, p. 90. 



Philoponus' Proofs of Creation 



115 



was presupposed. Aristotle had already demonstrated the finiteness of the body 
of the universe at considerable length, 150 and Philoponus did not trouble to take 
the subject up again. Still, completeness would require that the proof should 
include grounds for the finiteness of the body of the universe; Saadia's version 
of the proof did, for example, include such grounds. 151 The argument for the 
finiteness of bodies formulated by Kindi could serve the purpose well. It is an 
adaptation of one of Philoponus' other proofs for creation 152 and it could fittingly 
be adduced to support a key premise in Philoponus' proof for creation from the 
finite power of finite bodies. Kindi's sense that establishing the finiteness of bodies 
is pertinent to the issue of creation might then echo a version of the proof of 
creation from finite power in which the finiteness of bodies was explicitly 
demonstrated. 

The other detail that deserves attention is a sentence appearing at the end of 
Kindi's argument for the finiteness of bodies, in one of the passages in which he 
advances the argument. Apropos of nothing either preceding or following, Kindi 
asserts: "Actuality proceeds from power (al-fi'l kharij min al-quwa) inasmuch 
as the latter is the cause of the former; actuality [therefore] is finite by virtue of 
the finiteness of power." 153 In its context, the statement is completely incompre- 
hensible. It could, however, fit exactly into a proof of creation from the finite 
power of finite bodies. After establishing that the body of the universe is finite 
and finite bodies contain only finite power, such a proof might appropriately 
explain that "actuality," that is to say, actual existence, must in every instance 
proceed from a "power"; when actual existence proceeds from "finiteness of 
power" it must likewise be "finite"; and therefore the physical universe — con- 
taining, as it does, only finite power — can have existed for only a finite time. 

We have, then, two revealing details: Kindi's feeling that the finiteness of 
bodies is pertinent to proving creation; and the statement at the end of his argu- 
ment for the finiteness of bodies to the effect that the finiteness of "actuality" is 
a consequence of the "finiteness of power." Those two details could well be echoes 
of a lost version of Philoponus' proof for creation from the finiteness of the body 
of the universe. The conjectured version would establish the finiteness of the 
body of the universe by means of the principle that an infinite cannot be increased. 154 
It would proceed to establish the principle that a finite body can contain only 
finite power. It would contend that the "actuality" of a finite body is necessarily 
finite because of the "finiteness of [the] power" of the body. And it would con- 
clude that finite bodies, including the finite body of the universe, can have existed 
for only a finite time. 



150 Aristotle, Physics III, 5; De Caelo 1, 5-7. Cf. below, pp. 254-259. 

'"Above, n. 80. 

'"Above, p. 108. 

'"Kindi, Rasa'il. I, p. 196. 

" 4 Cf. above, p. 88; below, pp. 125-126. 



( 



116 Philoponus' Proofs of Creation 

( 

4. Summary 

Two sets of proofs for creation were formulated by Philoponus. One set rested 
' on the impossibility of an infinite number and argued in three different ways that 

, the number of past transformations and motions cannot be infinite. The other set 

rested on the principle that a finite body can contain only finite power and reasoned 
I that the finite body of the universe could not contain the power required for 

^ existence to have continued for an infinite past time. The latter set consisted of 

a basic proof supported by auxiliary arguments, among which were to be found 
( two arguments from composition and an argument from the continual succession 

( of forms across matter. Evidence from Arabic philosophers and bibliographers 

establishes that the works in which Philoponus developed his two sets of proofs 



( were known to the medieval Arabs. Moreover, Arabic texts explicitly cite, in 

Philoponus' name, the first proof from the former set and the basic proof of the 
latter. 

( Kindi and Saadia, who do not mention Philoponus by name, unmistakably 

employ some of Philoponus' proofs and may have derived other proofs from him. 
Kindi offers versions of the first and second proofs for creation from the impos- 
( sibility of an infinite number; and a remnant of the proof from the finite power 

I of finite bodies together with the auxiliary arguments from composition may be 

reflected in his treatment of the problem of creation. Saadia offers versions of all 
( three of Philoponus' proofs for creation from the impossibility of an infinite 

number as well as a simplified restatement of the proof for creation from the finite 
power of finite bodies. The auxiliary arguments from composition and the argu- 
( ment from the succession of forms across matter may also be reflected in Saadia. 

If my suggestion regarding the argument from the succession of forms across 
matter is correct, that argument was at some stage converted into a proof from 
( the succession of accidents in bodies. 

The next chapter will show that the three proofs of creation from the impos- 
' sibility of an infinite number were widely employed in the Middle Ages. The 

( proof from the presence of accidents in bodies also enjoyed considerable, although 

narrower, currency, for it was not taken seriously beyond the boundaries of the 
( Kalam. And several arguments from the composition in the universe were put 



( forward as well. 

( 
( 
( 
( 

I 
< 
( 
( 



i 



V 

Kalam Proofs for Creation 



1. Proofs from the impossibility of an infinite number 

As was seen in the previous chapter, John Philoponus drew up a set of three 
proofs for creation from the impossibility of an infinite number. All three were 
utilized by Saadia and two were utilized by Kindi, the form they took in Saadia 
and Kindi indicating that they had circulated in Arabic together with other proofs 
of Philoponus. The three proofs for creation from the impossibility of an infinite 
number enjoyed considerable vogue in medieval Islamic and Jewish philosophy, 
and they underwent several developments. 

Philoponus' first proof was the argument that the world must have a beginning 
since an infinite number of past transformations could not have been traversed. 1 
It became, in Kindi and Saadia, the argument that the world must have a beginning 
since an infinite past time could not have been traversed. 2 In the early period it 
was also employed by Iskafi (d. 854) and Nazzam (d. 845), two Mu'taziiite 
thinkers who were approximate contemporaries of Kindi. 

Iskafi is reported to have reasoned: "If there were no first |tcrml at which 
things begin, and which is not preceded by anything prior, it would be impossible 
for anything to come about." 3 The statement is not very precise, but we seem to 
have, once again, the familiar contention that an infinite series of events or an 
infinite continuous time — the text does not indicate which of the two Iskafi might 
have meant — cannot be traversed, and therefore, the present moment could never 
have been reached if such an infinite had preceded. Nazzam is similarly reported 
to have reasoned: "The traversal of bodies [sic] that has passed must be cither 
finite or infinite. ... If it were infinite, it would have no first [term], and what 
has no first [term] cannot be exhausted. The fact that what has passed has been 
exhausted is a proof of its finiteness." 4 At first glance one might suppose that 



1 Above, pp. 87-88. 

2 Above, pp. 95-96, 107. 

3 Khayyat, K. al-Inlisar. ed. and trans. A. Nader (Beirut, 1957), §5. 
"Ibid., §20. 



117 



( 

( 

(' 

118 Kalam Proofs for Creation 

{ 

. Nazzam is speaking of bodies' traversing space and that he is undertaking a proof 

of the spatial finiteness of the universe. But since his argument is represented in 

( our source, Khayyat, as a method for establishing creation, 5 and since it is dis- 

tinguished there from Nazzam's proof of the finiteness of the body of the uni- 
verse, 6 it must be interpreted in terms of the proof already known to us from 

I Kindi, Saadia, and now — with a bit of hesitation — from Iskafi: An infinite time, 

Nazzam is contending, could not be traversed by the succession of "bodies" that 
have existed; therefore, on the assumption of eternity, the past could never be 

' completed, and the present moment could never have been reached. 7 

We are told that a critic, Ibn al-Rawandi, accused Nazzam of self-contradiction, 
since by the side of the argument just cited he held that "what traverses and what 



( is traversed are infinite, and thus by maintaining that the former exhausts the 

latter Nazzam has affirmed the exhaustion of what is infinite." 8 The exact meaning 
^ of the objection is clarified by other information concerning Nazzam and by 

( recalling an objection that Saadia took up in connection with his fourth proof of 

creation. 9 Nazzam was notorious for having held the unorthodox position that 
^ bodies are infinitely divisible, at least in thought, a position implying that space 

( too is infinitely divisible. 10 Accordingly, Ibn al-Rawandi maintains, Nazzam him- 

self must assume the traversal of an infinite number of parts whenever a body 
^ moves from one place to another; why then should he deny the traversal of an 

( infinite time? This objection to Nazzam, which encapsules Zeno's first paradox, 



is more explicit in a parallel critique of Nazzam's proof of the finiteness of the 
K body of the universe." No clear solution to the objection is provided by our 

I ( source , Khayyat , but from other reports Nazzam is known to have held that bodies 

; succeed in moving from one place to another by "leaping" over some of the 

infinite parts of the distance traversed. 12 Saadia must have had knowledge of the 
'( exchange over Nazzam's proof, for he gives the same proof, raises the same 

objection, explicitly rejects the solution based on the theory of the "leap," and 
( offers an alternative solution instead. 13 

( 
( 

( 5 Ibid. 

"Ibid., §19. 

( 7 By "(he traversal of bodies," Nazzam might also conceivably mean: the total distance traveled 

by all bodies that have existed in the universe. Still another interpretation is given by II. Wolfson. 
( The Philosophy of the Kalam (Cambridge, Mass., 1976), p. 416. 

"Khayyat, K. al-lntisar, §20. 
/ "Above, p. 118. 

■ , '"References in A. Nader, Le Systeme Philosophique des Mu'tazila (Beirut, 1956), p. 155; H. 

' Wolfson, The Philosophy of the Kalam, pp. 495-496, 514. 

( "Khayyat, K. al-lnlisar, §19. For Zeno's paradoxes sec Aristotle, Physics VI, 9. 

l2 Refercnces in A. Nader, Le Systeme Philosophique des Mu'tazila. p. 182; H. Wolfson, The 
Philosophy of the Kalam, pp. 515-517. 

''Saadia, K. al-Amanat wa-l-I'tiqadat, cd. S. Landauer (Leiden, 1880), I, 1, 36-37; English 

( 
( 



Kalam Proofs for Creation 



119 



The impossibility of eternity was established by Nazzam in another way as 
well, described by Khayyat as one "of the best" the Kalam offered; "The stars 
cither arc equal to each other [ in velocity] ... or else some arc swifter than 
others. If they are equal |in velocity!, the distance traversed by some is less than 
what is traversed by all of them together, so that when the distance traversed by 
some is added to the distance traversed by the rest, the total will be greater than 
the part." Neither the total distance traveled by all the stars nor the distance 
traveled by a single star could, then, be infinite, for one infinite cannot conceiv- 
ably be greater or less than another. "If [the second alternative should be correct 
and) some stars arc swifter than others," the outcome is more obvious. For the 
distance traveled by the swifter is greater than the distance traveled by the slower, 
and "whatever is susceptible of less and greater is [as already seen] finite." That 
is to say, eternity would involve disparate infinites, whereas one infinite cannot 
be greater than another. Since the stars have traveled only a finite distance and 
hence can have performed only a finite number of revolutions, the heavens and 
the world as a whole must have existed for only a finite time. 14 By reasoning 
here from the principle that one infinite cannot be greater than another, Nazzam 
is giving a fairly faithful version of Philoponus' second proof of creation, the 
proof resting on that principle. The illustration employed by Nazzam, namely 
the movements of the stars, is also derived from Philoponus, 15 although Nazzam, 
unlike Philoponus, focuses on the distances that the stars travel rather than on 
the numbers of revolutions they perform. A new twist, moreover, has been added 
to the argument by Nazzam. Even if the stars happen to have identical velocities, 
he explains, the assumption of eternity still entails incommensurate infinites, 
since the distance traveled by each of the stars will be less than the total distance 
traveled by all. 

Philoponus' first proof for creation from the impossibility of an infinite number, 
the argument that the present could never have been reached if infinite past time 
or an infinite series of past events had to be traversed, has now been traced in 
Iskafi, Nazzam, Kindi, and Saadia. It appears with variations in Avicenna (who 



translation, with pagination of Arabic indicated: Book of Beliefs and Opinions, trans. S. Rosenblatt 
(New Haven, 1948). 

'"Khayyat, K. al-lntisar. §20. 

"in Philoponus' main statement of the proofs from the impossibility of an infinite number, the 
movements of the stars were the phenomenon upon which the third proof focuses; cf. above, p. 89. 
In a more casual statement of the proofs, the movements of the spheres, rather than of the stars, were 
among the phenomena adduced by Philoponus in the course of his second proof; cf. Commentary on 
Physics, ed. H. Vitclli, Commentaria in Aristotelem Graeca. Vol. XIV/2 (Berlin, 1897). p. 429. 
Also cf. Saadia as quoted above, p. 99. 



120 



Kalam Proofs for Creation 



rejects it), 16 Ibn Hazm. 17 Nasir-i-Khosraw, 18 Juwaynl, 19 Ghazali, 20 Judah Hal- 
levi, 21 Fakhr al-DIn al-RazI, 22 Averroes (who rejects it), 23 Maimonides (who also 
rejects it), 24 Amid!, 25 and Levi Gcrsonidcs. 26 It was known to the Christians, 
being accepted by Bonaventure 27 and rejected by Albertus Magnus 2 " and Thomas 
Aquinas. 29 Spinoza records it, 30 and it filtered down to Kant. 31 

Philoponus' second proof was the contention that the world must have a begin- 
ning since the past is continually being increased and an infinite cannot conceiv- 
ably be increased. The proof has now been discovered in Nazzam, Kindi, and 
Saadia. It appears as well in 'Abd al-Jabbar, 32 Avicenna (who rejects it) 33 lbn 
Hazm, 34 Ghazali, 35 Bahya, 36 Judah Hallevi, 37 ShahrastanI, 38 Ibn Tufayl (who is 



l6 In chapter 4 of an unpublished treatise contained in British Museum, MS. Add. Or. 7473, and 
summarized by S. Pines, "An Arabic Summary of a Lost Work of John Philoponus," Israel Oriental 
Studies, II (1972), 348. 

"Ibn Hazm, K. al-Fasl ft al-Milal (Cairo. 1964), I, p. 16 (proof 2); Spanish translation: Aben- 
h&zam de Cdrdoba y su Historia Crltica de las Ideas Religiosas, trans. M. Asm Palacios, Vol. II 
(Madrid, 1928), p. 102. 

l8 Cf. S. Pines, Beitraege zur Islamischen Atomenlehre (Berlin, 1936), p. 37, n. 2. 

"AT. al-Irshad (Cairo, 1950), p. 26; K. al-Shamil (Alexandria, 1969), p. 216; Textes apologe- 
tiques de Guwainl (Luma' ) ed. M. Allard (Beirut, 1968), p. 126. 

20 al-lqtisad ft al-l' tiqad (Ankara, 1962), p. 32; al-Risala al-Qudsiya, published as Al-Ghazali's 
Tract on Dogmatic Theology, cd. and trans. A. Tibawi (London, 1965), pp. 17, 35. 

2 'Kuzari, V, 18(1). 

22 K. al-Arba'in (Hyderabad, 1934), p. 15. This and Philoponus' other two objections to eternal 
motion are not given as complete proofs of creation by Razl. His complete argument is that if the 
universe were eternal it would have to be either eternally at rest or eternally in motion, but cannot 
in fact be either. 

"K. al-Kashf, ed. M. Mueller (Munich, 1859), p. 36; Tahafut al-Tahafut, cd. M. Bouygcs 
(Beirut, 1930), I, p. 20; English translation, with pagination of the Arabic indicated: Averroes' 
Taliafut al-Tahafut, trans. S. van den Bergh (London, 1954). 

"Guide to the Perplexed, I, 74 (2). 

"Ghdya al-Maram (Cairo, 1971), pp. 13-14. 

26 Milhamot ha-Shem (Leipzig, 1866), VI, i, 11, pp. 344-345. 

"Commentary on II Sentences, in Opera Omnia, Vol. II (Quaracchi, 1882), d. 1, p. 1, a. 1, 
q. 2; cf. E. Gilson, The Philosophy of St. Bonaventure (New York, 1938), p. 192. 
2e Physics, VIII, i, 12. 

29 Summa contra Gentiles, II, chap. 38 (3); Summa Theologiae, I, 46, art. 2, obj. 6. 
10 Cogitata Metaphysica, II, 10, 11. 

11 Critique of Pure Reason, A426/B454. For Mendelssohn's comments on the proof, cf. A. 
Altmann, "Moses Mendelssohn's Proofs for the Existence of God," Mendelssohn Studien, II (1975), 13- 
14. 

32 K. al-Majmu' ft al-Muhit bi-l-Taklif, cd. J. Houben (Beirut, 1965), p. 61. 
"Chapter 8 of the unpublished treatise cited above, n. 16. 
34 K. al-Fasl fl al-Milal, I, pp. 16-17 (proof 3). 
"al-Iqtisad, p. 33. 

M al-Hidaya (Robot ha-lebabot), ed. A. Yahuda (Leiden, 1912), I, 5, p. 44. 
"Kuzari, V, 18 (1). 

38 AT. Nihaya al-Iqdam, ed. A. Guillaume (Oxford and London, 1934), p. 28. 



Kalam Proofs for Creation 



121 



noncommittal regarding its validity), 39 Fakhr al-DIn al-RazI, 0 Maimonides (who 
rejects it), 41 TusI, 42 Levi Gersonidcs, 43 Ijl, 44 and Aaron ben Elijah. 45 It too was 
known to the Scholastics. Bonaventure accepted it, 4 '' Aquinas rejected it, 47 and 
its validity was hotly debated by subsequent Latin writers. 48 

The standard response of the medieval Aristotelians to the proof for creation 
from the impossibility of an infinite's being increased was that past time or past 
events arc not enumerable objects; consequently, the passage of infinite time 
would not after all involve the paradox of an actual infinite number's being 
increased. That response will be examined a little more fully below. A more 
radical response was also made, however. The more radical response counte- 
nances additions to, and subtractions from actual infinite numbers, but maintains 
that the notions greater, smaller, and equal are not applicable to the infinite. In 
other words, an infinite number can indeed be added to or subtracted from; but 
the resultant new infinite number cannot properly be described as greater than, 
smaller than, or even equal to the old infinite number. The paradoxes connected 
with incommensurate infinite numbers thereby fall away: Infinite numbers arc 
neither incommensurate nor commensurate. This response to the proof from the 
impossibility of increasing the infinite was made by Crescas 49 and certain Scho- 
lastic writers; 50 and a similar line of thought was later to be advanced by Bruno 
and Galileo. 51 As far as is known, only a single medieval Arabic writers, Thabit 
ibn Qurra, went so far as to countenance the incommensurability of infinites. 



m Hayy ben Yaqdhan, ed. and trans. L. Gauthier (Beirut, 1936), Arabic text, p. 81 , taken together 
with p. 76; French translation, p. 6 1 , taken together with p. 58; English translation: Hayy ben Yaqzan, 
trans. L. Goodman (New York, 1972), with pagination of the Arabic indicated. 

411 K. al-Arba'in, p. 15. Cf. above, n. 22. 

*' Guide, I, 74 (end). 

"Glosses to Razi's Muhassal (Cairo, 1905), p. 93. 
"Millmmot ha-Shem. VI. i, 11, pp. 341, 343, 346. 

44 K. al-Mawaqif (Cairo, 1907), VII, 224. lbn Tufayl, Tusi, and iji employ the method of "appli- 
cation" (cf. below, pp. 126-127) in formulating the argument. 
45 'Es Hayyim, ed. F. Delitzsch (Leipzig, 1841), p. 28. 

^Commentary on II Sentences, d. 1 , p. 1 , a. 1 , q. 2; cf. Gilson, The Philosophy of St. Bonaventure, 
p. 190. 

47 Summa contra Gentiles, II, 38 (4); cf. Summa Theologiae, I, 7. art. 4. 
4l< Cf. Maicr, "Problem des aktuell Uncndlichcn," Ausgehendes Mitlelalter, Vol. I (Rome, 1964), 
pp. 53-67, 76, 83. 

49 Or ha-Shem, III, i, 4. Crescas might have borrowed from the Scholastics; cf. S. Pines, ha- 
Skolaslika she-ahare Thomas Aquinas u-mishnalo shel Hasdai Crescas (Jerusalem. 1966). 
'"Maicr, "Problem des aktuell Unendlichen," as above, n. 48. 

51 G. Bruno, De la causa, V (beginning); Galileo, Discorsi e dimostrazioni matematiche intorno 
a due nuove scienze, in Opere (Florence, 1898), VIII, p. 78; English translation: Two New Sciences, 
trans. S. Drake (Madison, 1974), p. 40 (dealing with infinite spatial magnitude). Also cf. Spinoza, 
Ethics, I, xv, Schol.; H. Wolfson, Philosophy of Spinoza, Vol. I (Cambridge, Mass., 1948), pp. 288 
ff. 



122 



Kalam Proofs for Creation 



Thabit maintains that one infinite can in truth be "a third or a quarter or a fifth 
of another infinite." 52 

Philoponus' third proof of creation from the impossibility of an infinite number 
was the contention that since the planets move at different speeds, eternity would 
involve the absurdity of one infinite's being the multiple of another. The argument 
has been shown to appear with minor changes in Saadia. 53 It also appears in 
Avicenna (who rejects it), 54 in Ibn Hazm, 55 Ghazali, 56 Judah Hallevi, 57 Shah- 
rastanl, 58 Abu al-Barakat, 59 Fakhr al-DIn al-RazI, 60 Gersonides, 61 and a number 
of Scholastic writers. 62 Saturn, the planet that performs the fewest revolutions 
and hence gives rise to the greatest paradoxes, caught the fancy of the Muslim 
writers and is mentioned by most of them who offer the proof. 

Because of the similarity between the third proof and the second — one num- 
ber's being the exact multiple of another is simply a special case of its being 
larger than the other — the boundary between the two is generally blurred. Even 
Philoponus had not always kept the two distinct, doing so only in the main 
presentation of his proofs of creation. 63 With the exception of Saadia and Ghazali , 
all the instances of the third proof cited here conflate it with the second proof. 
The reasoning invariably is that since the revolutions of the planets are multiples 
of one another, eternity would involve the absurdity of one infinite's being larger 
than another. 

The train of thought animating proofs of creation from the impossibility of an 
infinite number was such that arguments could easily proliferate. Two obvious 
roads were open: First, additional objects could be discovered which would run 
to infinity on the assumption of eternity, and secondly, additional grounds could 
be discovered for the impossibility of an infinite number. 

The first road had already been taken by Philoponus. In his main presentation 
of the proofs from the impossibility of an infinite number, Philoponus had con- 
cerned himself with transformations and motions in the universe, explaining that 



52 Cf. S. Pines in Actes du Xle Congris International d'Hisloire des Sciences (Warsaw, 1968), 
III, p. 164. 

" Above, p. 99. 

'"Chapter 8 of the unpublished treatise cited above, n. 16. 
"K. al-Fasl ft al-Milal, I, p. 16 (3). 

56 al-Iqtisdd, p. 32; al-Risala al-Qudsiya, pp. 17, 35; Tahafut al-Falasifa, ed. M. Bouyges (Beirut, 
1927), I, §16; English translation in Averroes' Tahafut al-Tahafut, trans. S. van den Bergh (London, 
1954), p. 9. 

"Kuzari, V, 18(1). 

'*K. Nihaya al-lqdam, p. 29. 

59 K. al-Mu'tabar (Hyderabad, 1939), HI, 42. 

M K. al-Arba'm, p. 15. Cf. above, n. 22. 

61 Milhamot ha-Shem, VI, i, 11, p. 341. 

62 Maier, "Problem des aktuell Unendlichen," as above, n. 48. 

"Cf. above, pp. 88-89. 



Kalam Proofs for Creation 



123 



on the assumption of eternity, the number of transformations and motions occur- 
ring in the universe would run to infinity. But when the proofs appear elsewhere 
in Philoponus, additional illustrations are given. Philoponus points out that on 
the assumption of eternity, the members of the human species and indeed the 
members of each human genealogical line would run to infinity. The number "of 
horses," "of dogs," "of the other animals and plants as well as the movements of 
each of the spheres" would be infinite. There would accordingly result a "two- 
fold," a "threefold," a "many-fold, if not infinitely multiplied . . . actual infinite," 
all of which is ruled out by the impossibility of an infinite's being traversed, 
added to, or multiplied. 64 

The Arabic writers discovered a variety of things that would run to infinite on 
the assumption of eternity. Kindi, as was seen, contended that the segments of 
past time would run to infinity. 65 Saadia contended that the past time continuum 
would extend infinitely. 66 Ibn Hazm, probably by pure coincidence, adduces some 
of the examples that Philoponus had offered in his secondary statements of the 
proofs. Over an eternity, Ibn Hazm writes, the number of men, the number of 
horses, and the total of the two would be infinite, leading once again to the 
absurdity of one infinite's being larger than another. 67 Gersonides observes that 
the number of lunar eclipses would be infinite, which would mean that the moon 
is in an eternal state of eclipse! 68 Ghazali, followed by others, argues that an 
infinite number of immortal souls would accumulate, whereas even the strictest 
Aristotelian grants that an infinite number of things existing together is impossible. 69 

Ghazali goes a step further. Even if individual human immortality should be 
denied, he writes, souls or other objects can nonetheless be "supposed" to have 
come into existence at every moment of past time, to remain in existence, and 
to accumulate in an infinite number; yet an infinite number of objects is, for the 
various reasons that have been given, impossible. 70 No more is said by Ghazali, 
but his intent can be surmised. He undoubtedly was thinking of the Aristotelian 
definition, or characterization, of the "possible" as what is such that "when it is 



De Aeternitate Mundi contra Proclum, cd. H. Rabc (Leipzig, 1899), p. 11; Commentary on 
Physics, p. 429. 

""Above, p. 107. ""Above, pp. 95-96; 98-99. 

1,1 K. al-Fasl ft al-Milal, I, p. 16 (3). 

''"'Milhamot ha-Shem, VI, i, 11, pp. 342-343. This argument is descended from Philoponus ' 
third objection to eternal motion (cf. above, p. 89); for Gersonides is contending that it is absurd 
to suppose that a fraction of an infinite is equal to the whole infinite. 

69 7a/iu/M/ al-Falasifa, I, §22; English translation, p. 13. Also cf. Shahrastanl, K. Nihaya al- 
lqdam, pp. 24, 28, 50; Maimonides, Guide, I, 74 (7); Albertus Magnus, Physics, VIII, i, 12. in 
Opera Omnia, cd. A. Borgnct, Vol. Ill (Paris, 1890); Bonavcnturc, Commentary on II Sentences, 
d. 1, p. I, a. 1, q. 2; Gilson, The Philosophy of St. Bonaventure , p. 194; Aquinas, Summa contra 
Gentiles, II, chap. 38 (6), and Summa Theologiae, I, 46, art. 2, obj. 8; Maier, "Problem des aktuell 
Unendlichen," pp. 52-53, 56-57, 73, 76. Maimonides, Albertus, Aquinas, and Duns Scotus, all 
reject the argument; cf. H. Wolfson, The Philosophy of Kalam, pp. 455, 457-458, and Maier, ibid. 

70 Tahaful al-Falasifa, IV, §§8, 19; English translation, pp. 162, 169. Cf. below, p. 369. 



124 



Kalam Proofs for Creation 



assumed to exist, nothing impossible results therefrom." 71 His meaning is that 
even if objects have not actually come into existence at every moment of past 
time and remained in existence, for them to have done so is, nonetheless, a 
logical possibility. 72 Since it is possible, nothing impossible might result from 
assuming that it occurred. The assumption, however, that objects have at every 
moment of past time come into existence and remained in existence does result 
in an impossibility when combined with the additional assumption that the world 
is eternal; for the two assumptions taken together entail the existence of an infinite 
number of objects, something known to be impossible. Inasmuch as the former 
assumption, the "supposition" that an object has come into existence at each 
moment of past time, is possible, the latter assumption, the eternity of the world, 
must be rejected. 

Ghazali's contention that eternity would imply the actual accumulation of an 
infinite number of immortal souls or would at least permit "supposing" the accu- 
mulation of an infinite number of souls, played a central role in Scholastic spec- 
ulation regarding the validity of proofs from the impossibility of an infinite number. 73 

The first road to new arguments for creation from the impossibility of an infinite 
number consisted, then, in the discovery of additional objects that would run to 
infinity on the assumption of eternity. The second road to new arguments con- 
sisted in proposing additional reasons why past events, time, and the like, cannot 
be infinite. There was the contention that number is finite by its very nature; 74 
and consequently, the number of the revolutions of the spheres, 75 of past indi- 
viduals, 76 or of time itself, 77 must be finite. Alternatively, quantity was held to 
be finite by its very nature, so that time, which is a species of quantity, must be 
finite. 78 It was further maintained that whatever has a part must itself constitute 
a whole, and since a part of time can be marked off, the totality of time must 
constitute a whole and hence is delimited and finite. 79 Several writers laid down 
the principle that whatever has an end must have a beginning, whence the infer- 
ence was drawn that past time, ending as it does at the present moment, must 

71 Prior Analytics I, 13, 32a, 18-20. 

72 Ghazali does not take account of the rule that whatever is generated is eventually destroyed. If 
that rule should be regarded as a law of logic, generated objects would not be able to remain in 
existence eternally. 

"Gilson, The Philosophy of St. Bonaventure, p. 194; Maier, "Problem des aktucll Uncndlichcn," 
pp. 52-53, 56-57, 73, 76. A similar form of reasoning is employed by Aristotle, Physics VIII, 1, 
243a, 1-2; Metaphysics IX, 4. 

74 Cf. Aristotle, Physics III, 5, 204b, 7-10. 

73 Shahrastani, K. Nihaya al-Iqdam, p. 28. 

76 Judah Hallevi, Kuzari, V, 18 (1); Aaron ben Elijah, 'Es Hayyim, p. 28. 
77 Ibn Hazm, K. al-Faslfi al-Milal, I, p. 18 (4). 

78 Cf. Kindi's argument, above, p. 109; Gersonides, Milhamot ha-Shem, VI. i, 11, pp. 331-332. 

79 Ibn Hazm, K. al-Faslfi al-Milal, I, p. 15 (1); p. 17 (3); Bahya, al-Hidaya (Hobot ha-Lebabot), 
I, 5; Judah Hallevi, Kuzari, V, 18 (1). S. Pines, Beitraege zur Islamischen Atomenlehre , p. 15, n. 1 , 
connects the argument with Alexander of Aphrodisias. Also. cf. Euclid, Elements, V, definitions 1 
and 2. 



Kalam Proofs for Creation 



125 



have begun at some previous moment. Finally, the consideration was put for- 
ward that the number of the revolutions of the spheres must, like all numbers, be 
either odd or even', 81 whereas an infinite number could be neither. 82 A natural 
response to the last consideration was elicited from Crcscas, who maintained that 
the distinction of odd and even simply does not apply to infinite numbers. 83 

Thus far we have been examining medieval proofs of creation from the impos- 
sibility of an infinite number. At an early date the proofs were adapted by Arabic 
writers to serve fresh purposes, in particular to establish the finitcness of the 
body of the universe. 

Nu/./.am is reported to have refuted the Manichacans by contending that the 
denizens of darkness could never have crossed their own infinite realm to reach 
the realm of light, which they allegedly attack. For, Nazzam explains, "traversing 
an infinite is impossible, . . . and exhausting something shows it to be finite." 84 
Here we have the reasoning of Philoponus' first proof of creation 85 transferred 
to the subject of the body of the universe. Another consideration is also advanced 
by Nazzam: What is of finite extent in one direction must be so in all directions. 86 
That, apparently, is a utilization, vis a vis the body of the universe, of the principle 
that what has an end must have a beginning. 87 

In the previous chapter, Kindi was seen to have proved the finiteness of every 
body through a recasting of Philoponus' second proof of creation. Kindi's rea- 
soning was basically that magnitudes are subject to increase whereas the infinite 

80 Job of Edessa, Book of Treasures, trans. A. Mingana (Cambridge, England, 1935). I. v, p. 16; 
Abu al-Hudhayl, according to both Khayyat, K. al-lntisar, §5, and Shahrastani, K. al-Milal wa-l- 
Nihal, ed. W. Curcton (London, 1846), p. 35 (and cf. pp. 326-327); German translation of Shah- 
rastani, with pagination of the Arabic indicated: Religionspartheien und Philosophenschulcn, trans. 
T. Ilaarbrucckcr (Halle, 1850-1851); Maturidi, A'. al-Tawhld (Beirut, 1970), p. 14; Ibn Hazm. K. 
al-FasI ft al-Milal, I, pp. 18-19 (5); Bahya, al-Hidaya {Hobot ha-Lebabot), I, 5; Judah Hallevi, 
Kuzari, V, 18 (1); Avcrrocs, Tahafut al-Tahafut, I, p. 22. 

*'Cf. Aristotle, Categories 10, 12a, 7-8. 

82 Abu Bakr ibn Zakarlya RazI, Opera, cd. P. Kraus (Cairo, 1939), p. 129 (and cf. p. 130 f.); 
Ghazali, al-Iqtisad, p. 32; al-Risala al-Qudsiya, pp. 17, 35; Tahafut al-FalasiJ "a, I, §16; Gersonides, 
Milhamot ha-Shem, VI, i, 1 1 , p. 333. The argument seems to have been known to Thabit ibn Qurra; 
cf. S. Pines, above, n. 52. Avcrrocs agrees that an actual infinite number could be neither odd nor 
even and hence is impossible, but he docs not agree that past events can be assigned an actual number. 
Cf. Tahafut al -Tahafut, 1, p. 24; Epitome of Physics, \n Rasa' il Ibn fcii.v/i(/(IIydcrabad, 1947), III. pp. 
26-27, and Middle Commentary on Physics (Oxford, Bodleian Library, Hebrew MS. Ncubauer 
1380 = Hunt. 79), III, iii, 4, 2, p. 27b. 

"Or ha-Shem, 1, ii, 2; H. Wolfson, Crescas' Critique of Aristotle (Cambridge, Mass., 1929), 
pp. 219-221. Cf. Amidi, Ghaya al-Maram, p. 11. 

84 Khayyat, A", al-lntisar, § 19. Cf. passages cited by H. Bonitz, Index Arislotelicus (Berlin. 1 870), 
p. 74b, lines 30-34. 

"Cf. above, p. 87. 

* 6 K. al-lntisar, §19. The contention appears in Averroes and in Jewish writers dependent upon 
him; cf. H. Wolfson, Crescas, pp. 429-431. 
87 Cf. above, n. 80. 



126 



Kalam Proofs for Creation 



cannot be increased; hence neither magnitudes nor bodies, which arc a species 
of magnitude, can be infinite. 88 Similar argumentation appears in Baghdad!, 89 
'Abd al-Jabbar, 90 and Shahrastanl. 91 The same line of reasoning, that is, the 
recasting of Philoponus' second proof of creation to establish the finiteness of 
bodies, is to be found in Avicenna too, despite his rejection of the original use 
of the proof to establish creation. 92 One step in Avicenna's procedure is identical 
with the device, met in Kindi, of assuming that a segment is removed from the 
purported infinite. 93 Avicenna, however, incorporates added features that deserve 
separate treatment. 

Avicenna is, to be precise, demonstrating the impossibility of an infinite "con- 
tinuous quantity" of the type whose parts "exist together" and "have [relative] 
position," that is to say, the impossibility of an infinite line, plane, solid, or 
place. 94 He proceeds as follows: In order to demonstrate the impossibility of an 
infinite magnitude of the type specified, account need be taken only of magnitudes 
supposedly infinite at one end while finite at the other. As for magnitudes that 
are supposedly infinite at both ends, they can be assumed to be cut in the middle, 
so that the required finite end is provided. Given, then, a magnitude that is 
supposedly infinite at one end while finite at the other, the first step is to assume 
a segment removed from the finite end. The result, in effect, will be two mag- 
nitudes, each of which is finite at one end and infinite at the other, namely, the 
magnitude under consideration before the segment was removed and what remains 
of it after the segment is removed. The next step is to assume that the smaller of 
the two magnitudes is superimposed on, or "applied" to, the larger with the finite 
ends coinciding. The infinite ends could not now coincide; for should they coin- 
cide, the smaller magnitude would be equal to the larger, which is absurd. If, 
however, the infinite ends do not coincide, the smaller magnitude would be finite 
by virtue of being exceeded by the larger; and the larger would be finite by virtue 
of consisting of two finite magnitudes, to wit, the smaller magnitude and the 
segment that had been removed from the larger magnitude. The supposed infinite 
magnitude turns out to be finite, and an infinite magnitude is consequently 
impossible. 95 

The procedure just set forth is sometimes called the method of "application," 
since it involves applying one magnitude to another. 96 The method of application 
is employed to establish the finiteness of spatial magnitudes by a number of 



88 Above, p. 108. 

m K. Usulal-Din (Istanbul, 1928), p. 62. 

90 K. al-Majmu", p. 62. 

9, K. Nihaya al-Iqdam, pp. 24-25. 

92 Cf. below, p. 129. 

"Cf. above, p. 108. 

94 Cf. Aristotle, Categories 6, 51, 15 ff. 

"Avicenna, Najal (Cairo, 1938), p. 124. 

96 Cf. H. Wolfson, Crescas, p. 345. 



Kalam Proofs for Creation 



127 



philosophers who were dependent on Avicenna: Ghazali, in his compendium of 
Avicenna's philosophy; 97 Shahrastanl; 98 lbn Tufayl; 99 Abraham ibn Daud; 100 
Fakhr al-DIn al-RazI; 101 and Altabrizi. 102 

According to Avicenna, the method of application rules out not merely an 
infinite continuous quantity; it also rules out the existence of an "infinite . . . 
ordered number" of objects existing together. 103 Avicenna docs not spell out 
exactly how the argument is to run when directed against an infinite number of 
ordered objects. But Fakhr al-DIn al-RazI indicates what Avicenna intended 104 
when he, RazI, employs the method of application to disprove "an infinite regress 
of causes and effects [existing together]." RazI begins by assuming that several 
links arc removed from the supposed infinite series of causes. Then he imagines 
the new infinite series "applied" to the original scries with the finite ends coin- 
ciding. And he contends that the new infinite series could not conceivably extend 
back as far as the original scries, since it has fewer links. Yet it could also not, 
while remaining infinite, fall short. Consequently neither it nor the original scries 
could, after all, be infinite. 105 

Crescas, as will be recalled, rejected Philoponus' second proof of creation on 
the grounds that the notions smaller, greater, and equal do not pertain to the 
infinite. 106 On the same grounds, he rejects the use of the method of application 
to demonstrate the finiteness of the body of the universe. The question whether 
one infinite body applied to another infinite body would be smaller, greater, or 
equal, is, according to Crescas, illegitimate; for the concepts smaller, greater, 
and equal, just do not pertain to the infinite. 107 

2. Responses of the medieval Aristotelians to proofs of creation from the 
impossibility of an infinite number 

The proofs of creation which we have been examining presented a challenge 
to the medieval Aristotelians. The Aristotelians had to explain how (a) although 
Aristotle too had denied the possibility of an infinite number, 108 he could never- 
theless have advocated the eternity of the world, thereby implying the existence 
of infinite numbers of past objects and motions; how (b) although Aristotle had 

91 Maqcisid al-Falasifa (Cairo, n.d.), p. 127. 
™K. al-Milal wa-l-Nihal, p. 403. 

m Hayy ben Yaqdlum, Arabic text, p. 76; French translation, p. 58. 
im Emuna Rama (Frankfort, 1852), 1, 4, pp. 15-16. 
""K. al-Arba'm, p. 28. 

102 As cited by H. Wolfson, Crescas. pp. 346-347. 
ws Najal, pp. 124-125. 

"' 4 Mainionidcs also understood Avicenna's intent; cf. Guide, I, 73 (11). 

" I5 AT. al-Arba'm, p. 83. AmidT, Ghaya al-Maram, pp. 9-10. cites, but does not accept, the 
argument. 

"'"Above, p. 121. 

""Orha-Shem, I. ii, 1(a); Wolfson. Crescas, pp. 188-191. 

'""Physics 111. 5, 204b, 8-10; the reason Aristotle gives is that the infinite cannot be traversed. 



128 Kalam Proofs for Creation 

( denied the possibility of an infinite magnitude, 109 he could have affirmed infinite 

past time; and how (c) although he had denied the possibility of an infinite scries 
( of causes and effects existing together, ' 10 he could yet have affirmed the existence 

of an infinite series of causes and effects that succeed one another through time 
and do not exist together. 
( From Maimonides and Averroes we learn that the first two of the three diffi- 

/ culties were dealt with by Alfarabi in a work, now lost, entitled On Changeable 

Beings (Ft al-Mawjudat al-Mutaghayyira). Maimonides' report refers specifically 
I to the problem of infinite past individuals and motions . These , Alfarabi explained , 

continue to exist only "in imagination" ; since they do not exist together in actual- 
( ity, they cannot properly be enumerated and therefore are not affected by the 

( absurdity of an actual infinite number. 1 1 1 According to Averroes' report, Alfarabi 

offered a similar solution to the problem of infinite time, as distinct from the 
1 succession of past individuals and motions, which are the subject of Maimonides' 

( account. Some err, Averroes writes, by supposing that time cannot run to infinity 

just as a straight line cannot. In fact, Alfarabi explained, a straight line cannot 
' extend infinitely only because it "possesses position and exists in actuality"; and 

( since time does not resemble a straight line in either respect," 2 the grounds for 

rejecting an infinite straight line are absent in the case of time." 3 
C The reports of Maimonides' and Averroes', taken in combination, indicate that 

/ arguments against infinite numbers of objects or an infinite extension arc opera- 

tive, in Alfarabi's view, only when two conditions are met: The objects in question 
( and the parts of the purported infinite extension must exist together in actuality; 

I and they must possess position. Averroes and Maimonides do not reveal, how- 

ever, precisely which arguments against the infinite Alfarabi wished to restrict 
( through the two conditions nor what the import of the two conditions is. Sub- 

^ sequently, Avicenna offers an analysis similar to, but more nuanced than, Alfar- 

abi's, and he does explain the import of the conditions. Avicenna connects them 
C with the method of application, which he had employed to rule out certain types 

( of infinite. 

The method of application consisted in applying or superimposing one object 
' or one series upon another so that the parts of the one match, and are paired off 

( with, the parts of the other. Avicenna accordingly maintains that the method can 

be utilized only when the parts of each object or each series, firstly, "exist . . . 
( together," and secondly, either occupy "[relative] position" or are "essentially 

• ordered," essential order being the situation wherein the parts "precede one another 

( m Physics III, 5, 204b, 1 ff. 

"°Bclow, p. 337. 

'"Guide, I, 74 (end), supplemented by I, 73 (1 1), where Alfarabi is not mentioned. 
I " 2 As pointed out by Aristotle, Categories 6, 5a, 26-28. 

'"Epitome of Metaphysics, ed. and trans. C. Quirfjs Rodriguez (Madrid, 1919), IV, §4; German 
I translation: Die Epitome der Metaphysik des Averroes, trans. S. van den Bergh (Leiden, 1924), 

pp. 106-107. 



Kalam Proofs for Creation 



129 



naturally." Only when both conditions are met, clearly, can one object or scries 
be "applied" to another. The method of application would therefore rule out an 
infinite spatial magnitude, inasmuch as the parts of spatial magnitudes all exist 
together and possess relative position. The method would likewise rule out an 
infinite number of objects existing together in instances where the objects have 
an order in nature, for example, when they stand in the relation of cause to effect. 
But the method would not rule out infinite time and an infinite series of past 
motions, nor again an infinite number of unarranged objects existing together. 
The method would fail to rule out infinite time or an infinite number of past 
motions since neither the parts of time nor the several motions, although arranged 
in an order, exist together, and one series consequently cannot be applied to 
another. The method would fail as well to rule out an infinite number of unar- 
ranged objects that do exist together, such as an infinite number of "angels," "evil 
spirits," and the like. For entities of the sort, although existing together, do not 
satisfy the second condition; they do not possess relative position and one series 
cannot be applied to another in a manner that will pair the members off. 1 14 

Such, then, is Avicenna's explanation why an infinite extended object or an 
infinite scries of objects is impossible when the parts of the object and the links 
in the series satisfy both the condition of existing together and the condition of 
being arranged in order; and why infinite extension and an infinite series remain 
possible as long as cither of the requisite conditions is not fulfilled. Avicenna's 
explanation implies a judgment regarding arguments from the impossibility of an 
infinite number. The argument that an infinite series of past events or infinite past 
time cannot be traversed is, by implication, dismissed." 5 The sole argument 
from the impossibility of an infinite number which he takes seriously is the 
argument from the impossibility of an infinite number's being increased; and he 
understands it to be valid solely when formulated with the aid of the method of 
application, solely in situations where sets of actual objects can be superimposed 
upon one another. 

Ghazali, as was seen a little earlier, developed a version of the proof for creation 
from the impossibility of an infinite number wherein he contended that eternity 
would entail the accumulation of an infinite number of immortal souls, or at least 
the possibility of "supposing" the accumulation of an infinite number of souls 
and other objects." 6 Ghazali proceeds to explain — although without mentioning 
the method of application — that the infinite number of immortal souls and the 
like which would have come into existence, or which might be supposed to have 
come into existence, through infinite past time would satisfy the two conditions 
we have been considering. Immortal souls would exist together and, since they 

""Najat. pp. 124-125. 

1 "The impossibility of traversing an infinite is dealt with in chap. 9 of the unpublished work of 
Avicenna's summarizxd by Pines, "An Arabic Summary of a lost work of Philoponus," p. 349. 
" 6 Above,p. 123. 



130 



Kalam Proofs for Creation 



come into existence successively through time, would be arranged in order. The 
doctrine of eternity thus entails an infinite number of objects that meet the two 
conditions set by Alfarabi and Avicenna for the impossibility of an infinite num- 
ber. Ghazali concludes therefore that even granting the two conditions, the doc- 
trine of eternity entails an impossibility and is untenable." 7 

The development we have been following merits recapitulation. The Peripatetic 
principle that an infinite cannot be exceeded underlies one of Philoponus' proofs 
of creation: Philoponus reasoned that since the past is continually being increased, 
it cannot be infinite, and consequently the world must have a beginning. In the 
Middle Ages, Philoponus' proof of creation was readapted to prove the impos- 
sibility of an infinite magnitude and even the impossibility of an infinite series 
of causes. Medieval Aristotelians endorsed the adaptations of Philoponus' proof, 
while rejecting its original use to rule out an infinite past. The apparent discrep- 
ancy was dealt with by Alfarabi and Avicenna. They maintained that infinite 
objects and infinite series are excluded only when they exist together and possess 
an order, whereas infinite past time and the infinite series of objects entailed by 
infinite past time do not meet the two conditions. Ghazali thereupon answers 
Alfarabi and Avicenna. He points out that the assumption of infinite past time 
does entail an infinite number of objects satisfying the two conditions determined 
by them; the doctrine of eternity is therefore refuted from the standpoint of the 
Aristotelians themselves. 

We have seen how a single solution proposed by Alfarabi and Avicenna addressed 
two difficulties: the discrepancy in Aristotle's rejecting the possibility of an infi- 
nite number, while advocating the eternity of the world with its implication of 
infinite past events and objects; and the discrepancy in his rejecting infinite mag- 
nitudes while affirming the infinity of past time. The third difficulty that con- 
fronted medieval Aristotelians was the discrepancy between Aristotle's rejection 
of an infinite regress of causes, on the one hand, and a further implication of the 
eternity of the world, on the other. 

The eternity of the world, at least of an Aristotelian, noncvolutionary world, 
implies an infinite succession of generations, let us say, of fathers and sons. When 
the series of fathers and sons is thought of merely as an infinite number of objects, 
it has to be harmonized with arguments ruling out an infinite number, and we 
have seen how the harmonization might be accomplished. Quite apart, however, 
from being thought of merely as an infinite number of objects, past generations 
can also be thought of as an infinite causal series. Each father is the cause of the 
existence of his son; and over an eternity, every series of fathers and sons would 
constitute what might be termed an infinite diachronic causal regress. By advo- 
cating the eternity of the world, Aristotle and his followers advocated a diachronic 
causal regress. Yet he and his followers scrupulously rejected a synchronic causal 
regress, that is, an infinite regress of causes existing together. 

" 7 Tahafut al-Tahaful, IV, §19; English translation, p. 169. The point is further discussed, below, 
pp. 367-370. Amidl, Ghaya al-Maram, p. 1 1 , repeats Ghazali's criticism of the Aristotelians. 



Kalam Proofs for Creation 



131 



Various reasons were given by various philosophers for the impossibility of an 
infinite regress of causes existing together. 118 Aristotle's reason was that in a 
causal series, no link which is activated by something standing behind it can be 
deemed a true cause. The true cause of whatever occurs in a causal scries is the 
cause standing behind and activating the entire scries, that is to say, the first 
cause. Consequently, if there should be no first cause, there is no true cause, 
hence "no cause whatsoever," which is absurd." 9 Such being Aristotle's reason 
for rejecting an infinite causal regress, the problem that poses itself is this: An 
infinite regress of causes succeeding one another through time would seem to be 
excluded on the same grounds that Aristotle adduced to rule out an infinite regress 
of causes existing together: If there were no first cause, there would be no true 
cause, and hence no cause whatsoever. How then can Aristotle and his school 
espouse the eternity of the world, with the attendant infinite regress of causes 
extending back through time? 

Avcrrocs explicitly takes up the problem and he solves it with the aid of a 
distinction — alluded to by Maimonides as well 120 — between an accidental and 
an essential scries. An essential scries, as defined by Avcrrocs, is any scries in 
which the prior links arc an indispensable "condition" for the existence of the 
posterior links — prior and posterior being taken here in a broad sense that com- 
prehends both the temporal and nontcmporal. 121 An accidental scries is any scries 
in which the prior links arc not an indispensable condition for the existence of 
the posterior. Accordingly, an accidental series will not merely be one in which 
the links form an obviously noncausal succession. It will also be one in which 
the "prior is thought to be a cause of the posterior," if in reality the prior link is 
not a necessary condition of the posterior link. 122 The distinction between essen- 
tial and accidental series is illustrated by Averroes through the factors in the 
production of a given man. Although the generations of men seem to be connected 
causally, in fact "the existence of the prior [man] ... is not a condition of the 
existence of the posterior." 123 The scries of progenitors standing behind a given 
man is not a necessary condition for his existing, because the same man could 
have existed with other progenitors. The necessary and essential conditions of 
the man's existence, the genuine causes, are the permanent cosmic forces respon- 
sible for everything occurring in the universe. 124 

Now, Averroes explains, whereas an essential series, that is to say, a scries of 
genuine causes, cannot regress infinitely, an accidental series may do so, even in 
instances where the links appear to be connected causally. Avcrrocs might have 

""Above, p. 127 (the method of application); below, p. 339. 
u9 Metaphysics II, 2, 994a, 1-19; cf. below, p. 337. 
' 2 "Guide, I, 73 (11). 
m Cf. below, p. 338. 

' 22 Tahaful al-Tahaful, I, pp. 20-21. Cf. Epitome of Physics, p. 1 10; Long Commentary on Physics, 
in Arislotelis Opera cum Averrois Commentariis, Vol. IV (Venice, 1562), VIII, comm. 15. 
'"Tahafut al-Tahafut, IV, pp. 268-269. Cf. Long Commentary on Physics, VIII. comm. 47. 
'"Tahafut al-Tuhafut, I, p. 59. 



132 



Kalam Proofs for Creation 



put the proposition in the following way, although I did not find that he explicitly 
does: 125 Aristotle had rested the impossibility of an infinite regress of causes on 
the consideration that the true cause of a series is the first cause alone, and without 
a first cause the series would have no true cause. The reasoning clearly pertains 
solely to an essential series, to causes that are indispensable conditions of what 
ensues; for only among them is the true cause to be sought. Accidental series, 
which do not contain indispensable conditions for what ensues, would in no way 
be prevented from running to infinity. 

Having laid down the foregoing distinction between essential scries and acci- 
dental series, Averroes can maintain that the members of the essential series and 
only they must "lead upward to an eternal first cause." 126 The first essential cause, 
which is ultimately responsible for everything in the universe, is "the sphere, or 
the soul [of the sphere], or the intellect [of the sphere], or all together, or " — 
the preferable alternative — "the creator." 127 It, rather than any of the progenitors 
of a given individual, is the true cause of the appearance, as well as the continued 
existence, of the individual. 128 

As Averroes works out the distinction between accidental and essential series, 
he adds a further refinement or clarification. Fathers are, after all, observed to 
play a role in producing their offspring, and, what undoubtedly is at least as 
important, Aristotle had explicitly stated that a man is begotten by another "man 

'"Cf. Tahafut al-Tahafut, I, p. 22. 
^Tahafut al-Tahafut, IV, pp. 268-269. 

127 Ibid. By "creator," Averroes of course means the eternal cause of the existence of the universe. 

128 In several places Averroes connects the distinction between essential and accidental series with 
another distinction, that between a rectilinear series and a circular series. Unhappily, he gives differing 
accounts of the connection. In the Tahafut, he writes that series formed by essential causes are 
rectilinear, that is to say, the order is irreversible and no link in the chain of causes and effects can 
ever recur; and series formed by accidental causes are circular. The progenitors of an individual man 
and the stages in the rain cycle, both of which constitute accidental series, are, in the passage in 
question, both represented as being circular series. They are circular series because the amount of 
matter in the universe is finite and the processes of nature can be sustained only if material objects 
decay and the matter is reused — only if, in contemporary parlance, the matter is 'recycled.' To 
maintain the processes of life, progenitors must continually die, decay, and be transformed into 
nourishment for subsequent generations. To maintain the rain cycle, a given particle of water must 
turn to vapor, precipitate, fall back to earth as rain, and thereupon repeat the process. Cf. Tahafut 
al-Tahafut. I, pp. 56-57; IV, pp. 268 - 269, 274. 

In K. al-Kashf, p. 37, Averroes characterizes the rain cycle as a "circular" series, but now he 
characterizes the generations of fathers and sons as a "linear" and "accidental" series. Thus, in 
contrast to what he writes in the Tahafut, he recognizes series that are accidental as well as linear, 
the generations of men being represented as such. 

A resolution of the discrepancy might be proposed with the aid of Aristotle. De Generatione 
II, 11. There Aristotle writes that series such as the generations of man "appear" to be rectilinear, 
but in fact are so only "numerically," whereas in respect to "species" they too are circular. It might 
be tempting to interpret Averroes as follows: In the Taliafut, he means that all accidental series are 
circular in respect to species. That is to say, the same matter continually recurs in beings that are 
specifically, although perhaps not individually, the same; for example, the matter of a previous man 



Kalam Proofs for Creation 



133 



. . . together with the sun." 129 In one passage, therefore, Averroes is led to 
explain that the immediate progenitor is indeed indispensable for a given man's 
coming into existence, but solely as what Averroes now calls an instrument, and 
not as a cause in the strict sense: The immediate progenitor is an indispensable 
instrument for the action of the cosmic powers, which are the genuine causes. 
According to the present passage, it is the earlier progenitors that are purely 
accidental factors in the production of a particular man, and they, being acciden- 
tal, can — or must — regress infinitely. 130 

The question we have been considering is how Aristotle's rejection of an infinite 
regress of causes existing together could be harmonized with his espousal of the 
eternity of the world and the attendant infinite regress of causes extending back 
through time. Averroes' solution is that the former is an essential series whereas 
the latter is an accidental series, and the impossibility of an infinite regress covers 
only essential series, not accidental series. The distinction between an essential 
and an accidental series, together with Averroes' differentiation between the 
immediate instrument and prior instruments, is repeated by Thomas Aquinas. 131 

Resume 

The present chapter has thus far examined the history, in the Middle Ages, of 
Philoponus' three proofs of creation from the impossibility of an infinite number. 
The proofs were employed in their own right. And new versions were developed 
both through the proposing of additional objects that would run to infinity in an 
eternal universe and through the proposing of additional reasons, alongside Phil- 
oponus' three, for the impossibility of an infinite of one sort or another. Philo- 
ponus' proofs of creation, especially the proof turning on the impossibility of an 
infinite number's being increased, were also recast to serve fresh purposes, spe- 
cifically, to rule out an infinite spatial magnitude and an infinite series of causes. 
Even philosophers of the Aristotelian school, for whom the eternity of the uni- 
verse was a virtual dogma, employed Philoponus' argumentation for those other 



reappears in a later man. And in K. al-Kashf, Averroes means that the rain cycle is a case of a scries 
circular even in respect to its individuals, inasmuch as exactly the same drop of rain can recur; 
whereas the generations of men are linear in respect to their individuals, since exactly the same 
individual man cannot ever recur. This proposed interpretation, it must be confessed, docs not seem 
to harmonize with Averroes' remarks on circular scries at the end of his Epitome of the De Gener- 
atione. Sec the Arabic text of the Epitome of De Generatione, in Rasa' U Ibn Rushd (Hyderabad, 
1947), pp. 32-33; Hebrew text: Commenlarium Medium et Epitome in Aristotelis de Generatione et 
Corruptione Libros, Textum Hehraicum, ed. S. Kurland (Cambridge, Mass., 1958), pp. 125-126; 
English translation: Averroes' Middle Commentary and Epitome on Aristotle's De Generatione, trans. 
S. Kurland (Cambridge, Mass., 1958), p. 136. 

,29 Physics II, 2, 194b, 13, cited by Averroes, Tahafut al-Tahafut, IV, p. 268. Cf. Aristotle, 
Metaphysics XII, 5, 1071a, 13-17. 

""Tahafut al-Tahafut, IV, p. 269. Cf. Epitome of De Generatione, pp. 32-33;/C. al-Kashf. p. 37. 
It is not clear how if c is essential for b and b is essental for a, c is not essential for a. 

,}] Summa contra Gentiles. II, chap. 38; Summa Theologiae, I, 46, art. 2, reply to obj. 7. 



134 



Kalam Proofs for Creation 



purposes. Since the Aristotelians endorsed the utilization of Philoponus' argu- 
mentation for certain purposes, and since Aristotle too had ruled out certain types 
of infinite, the Aristotelians were faced with a problem of consistency. They had 
to explain why the grounds for ruling out some infinites do not affect other 
infinites that are implied by the doctrine of eternity. The requisite harmonizations 
were forthcoming: The grounds for the impossibility of an infinite number were 
shown to be limited in scope and not to cover the types of infinite implied by the 
doctrine of eternity. Isolated philosophers, wc have also seen, took a more radical 
position and rejected the various grounds for the impossibility of the infinite. 
They held that infinite time, infinite past events, and infinite spatial magnitudes, 
are all equally conceivable. 132 

3. The standard Kalam proof for creation: the proof from accidents 

Proofs for creation from the impossibility of an infinite number had a wide 
currency; but it was another proof , the one Saadia entitled "from accidents," that 
became the Kalam demonstration par excellence of creation. Briefly, the proof 
from accidents runs as follows: Since accidents are necessary concomitants of 
bodies and are subject to generation, bodies too must be subject to generation; 
the universe, which is a body, must therefore have been generated. In the previous 
chapter, I suggested that the proof may be an outgrowth of the second of the 
auxiliary arguments with which Philoponus had supported his overall proof of 
creation from the finite power of finite bodies. Philoponus had in that second 
auxiliary argument focused on the continual succession of forms across matter, 
whereas the proof from accidents focuses on the continued presence of acci- 
dents — the Kalam analogue of the Aristotelian forms— in bodies. 133 

The first thinker to offer the proof from accidents is reported to have been the 
early Mu'tazilite, Abu al-Hudhayl (d. 849). 134 The proof would accordingly go 
back to the first half of the ninth century. At the turn of the tenth century the 
proof is found in Saadia, and his version was examined in the previous chapter. 135 
The proof was also known to Alfarabi, a contemporary of Saadia's. Alfarabi, a 
dyed-in-the-wool Aristotelian, could not by any means have accepted the con- 
clusion, 136 but he cites the proof in one of his logical works as an example of a 
"compound syllogism." In order to add weight to the illustration, Alfarabi spins 
out the steps, so that the argument runs: "|1J Every body is composite. |2] 
Everything composite is joined to, and cannot be free of an accident [the accident 
of composition]. [3] Everything joined to, and not free of an accident is joined 
to, and not free of what is generated. [4] Everything joined to, and not free of 

m On the last point, see above, pp. 121-122, 127. 
m Cf. above, p. 92. 
l34 Cf. below, n. 162. 
135 Above, pp. 103-104. 

l36 He undoubtedly rejected the argument for the reason that Averroes gives for rejecting it; cf. 
below, pp. 143-144. 



Kalam Proofs for Creation 



135 



what is generated does not precede what is generated. [5] Everything that docs 
not precede what is generated has its existence together with the existence of 
what is generated. [6| Everything having its existence together with the existence 
of what is generated has its existence after nonexistence. |7] And everything 
having its existence after nonexistence is generated. But the world is a body. 
Consequently, the world is generated." 137 

The proof from accidents appears in Ibn Suwar, who explicitly labels it a Kalam 
method of proving creation, points to several flaws in it, and thereupon rejects it 
in favor of what he identifies as John Philoponus' proof from the finite power of 
bodies. 138 Ibn Suwar presents the proof in two versions, which he characterizes 
as reformulations in "a technical arrangement" of the original Kalam argument. 
What he means is that the original Kalam reasoning had been recast — by him or 
by someone before him— in syllogistic form. One of Ibn Suwar's syllogisms 
reads: "Body is not free of accidents nor does it precede them; but whatever is 
not free of accidents and does not precede them is itself generated; therefore body 
is generated." Ibn Suwar's other syllogism employs the term "generated things" 
in place of the term "accidents," and reads: "Body is not free of generated things 
|etc.]." 139 

The argument from accidents, then, reportedly goes back to Abu al-Hudhayl. 
It was advanced with approval by Saadia, cited undoubtedly without approval by 
Alfarabi, and cited with explicit disapproval by Ibn Suwar. Ash'ari also knew of 
the proof, employing it in one work, 140 but rejecting it in another. In the work 
where he rejects it, his reason is not that— as Alfarabi and Ibn Suwar thought — 
the proof is insufficiently philosophical but, on the contrary, that it is too phil- 
osophical. He ascribes the proof to "the philosophers, and those who follow them 
among the qadanya [i.e., the Mu'tazilites], the innovators [or: heretics], and the 
deviators from the prophct(s)." And he takes the position that the testimony of 
Scripture is more than adequate and no rational proof of creation is needed. 141 

Despite Ash'ari's reservations, later adherents of his school embraced the proof 
from accidents. The Ash'arite BaqillanI reasons: Everything in both the higher 



137 Epitome of Prior Analytics, cd. M. Turker. Revue de la Faculle de Lnngues, d'Histoire el de 
Geographic ile U University d' Ankara. XVI/3-4 (1958), 263. On p. 262, Alfarabi spells out each 
of the individual constituent syllogisms in full. Cf. H. Davidson, "John Philoponus as a Source of 
Medieval Islamic and Jewish Proofs of Creation," Journal of the American Oriental Society, LXXXIX 
(1969). 383. 

""Cf. above, p. 90. 

l19 Thc text is published in A. Badawi. ed., Neoplatonici aputl Arabes (Cairo. 1955). pp. 243, 
245; French translation in B. lxwin, "La Notion dc muhdat dans le kalam ct dans la philosophic," 
Orientalia Stienica, III (1954). 88-93. 

I4 "A'. al-Lttma', in The Theology of al-Ash'ari, ed. and trans. R. McCarthy (Beirut. 1953), S§6, 

93. 

""Risdlaildahlal-Tlmghr, Publications of the Theological Faculty, Istanbul, V1IK1928). 89; cf. 
R. Frank, "Al-Ash'arl's Conception of the Nature and Role of Speculative Reasoning in Theology," 
Proceedings of the Villi Congress of Arabic and Islamic Studies (Stockholm, 1972), pp. 138-141 . 



( 



( 



136 Kalam Proofs for Creation 



i 
( 
/ 

f and lower parts of the universe consists of "substances" (i.e. , atoms) and "acci- 

dents." Accidents are generated; that is shown by "the fact that motion is destroyed 
with the advent of rest, for it if were not destroyed with the advent of rest, both 
would be present together in the body, . . . something necessarily known to be 
impossible." BaqillanI might be expected hereupon to argue that not merely acci- 
dents but atoms too. are generated. Yet he does not do so, and proceeds instead 
to establish that "bodies" are generated. Bodies, he explains, "do not precede 
generated things nor exist before them," inasmuch as "bodies cannot avoid having 
their parts touching and joined, or separated." But anything "that does not precede 
what is generated is likewise generated," since it must either come into existence 
"together" with or "after" its generated concomitant. Both accidents and bodies 
are, then, generated, and the reader is left to conclude that since the universe 
consists only of accidents and of bodies, which arc conglomerations of atoms, 
the entire universe must be generated. 142 
( The versions of Saadia, Alfarabi, Ash'ari, Ibn Suwar, and BaqillanI reveal 

■ significant similarities. 143 All undertake to demonstrate the generation of bodies 

and accidents. All employ the proposition that bodies "cannot avoid" (Saadia, 
( BaqillanI: la yakhlu) or cannot be "free of" (Alfarabi, Ibn Suwar: la yanfakk; 

Ash'ari: lam yanfakk) their association with accidents. And all conclude that 
bodies are generated since they "do not precede" (Saadia, BaqillanI: lam tasbuq; 
'( Ash'ari: lam yasbuq; Alfarabi: ghayr sdbiq; Ibn Suwar: la yataqaddam) what is 

generated. In Saadia, Ibn Suwar, 144 and BaqillanI, the prime illustration of an 
accident is motion. 145 Alfarabi and BaqillanI adduce the accident of the com- 
I position or joining of bodies to establish the proposition that bodies are always 

associated with accidents. Saadia and BaqillanI seek comprehensiveness by tak- 
' i'ng into account the higher as well as the lower parts of the universe. They both 

( conclude in similar phrases that what does not precede generated things is "of 

the same character" (Saadia: mithluhu) and "likewise" (BaqillanI: kahuwa ) 
^ generated. 

( The Kalam doctrine of accidents is integral to the proof. Yet atomism, which 

is commonly thought to be a correlate of the Kalam doctrine of accidents, is 

( curiously absent, the preserved early versions establishing the generation of bod- 

( ies, not atoms. As a result, these versions do not prove creation ex nihilo: They 

establish the creation of the body of the universe, but not the creation of the 

( atoms from which the universe is constituted. The failure to prove the creation 

/ 

'* 2 K. al-Tamhid, ed. R. McCarthy (Beirut, 1957), pp. 22-23. 
( l43 Ash'ari's version, K. al-Luma', §§6, 93, is less complete than the others. 



As Ibn Suwar records the proof, the principle that no body is free of an accident is established 
through an analysis of the accidents of motion and rest. Cf. above, n. 139, Arabic text, p. 244; 
/ French translation, §3. 

""Ash'ari, K. al-Luma', also has the example of the accident of motion. It should further be 
/ recalled that Kindi, in another context, employs the proposition that body does not precede the 

accident of motion. Cf. above, p. 110. 



i 



Kalam Proofs for Creation 



137 



of atoms is especially conspicuous in Baqillanl's version. In a preface, BaqillanI 
had divided physical beings into bodies, atoms, and accidents; 146 and he opens 
his argument with the statement that the upper and lower regions of the universe 
arc composed of atoms and accidents. 147 The reader awaits a proof of the gen- 
eration ol accidents and atoms, or of accidents, atoms, and bodies, hence a proof 
of the creation of all physical existence ex nihilo. What BaqillanI in fact provides 
is a proof that accidents and bodies are generated. Such must have been the object 
of the ninth-century proof, which underlies the preserved early versions. 

Although the Kalam complexion of the proof is unmistakable and although 
from the beginning the proof was regarded as the property of the Kalam school, 
Baqillanl's version alone, among the versions examined thus far, is worked out 
within a Kalam conceptual frame. Saadia took the existence of accidents for 
granted, and relied on simple observation as grounds for the propositions that 
accidents arc subject to generation and that bodies are never free of them. For 
Alfarabi and Ibn Suwar, those points — the existence of accidents, the proposi- 
tion that accidents are subject to generation, the proposition that bodies are never 
devoid of accidents — were of minor importance, since Alfarabi and Ibn Suwar 
rejected the proof in any event. Ash'ari's version is incomplete. But BaqillanI 
has a standard analysis of each point. 

In his preface to the proof — which 1 have not quoted yet — BaqillanI establishes 
the existence of accidents by considering motion, showing that it cannot be due 
to the moving body "itself" — were that the case a moving body would never 
stop moving — but must rather be an added entity, a "something" (ma'nd), in the 
moving body. 148 This is the characteristic Kalam theory that accidents are actual 
entities inhering in atoms and bodies. 149 And the theory appears in BaqillanI in 
association with the related and very peculiar Kalam notion that rest is not merely 
a privation, that is to say, the absence of the accident of motion, but a positive 
quality and, no less than motion, a "something." 150 

When BaqillanI proceeds, within the proof proper, to establish the proposition 
that accidents are generated, he adduces "the fact that motion is destroyed with 
the advent of rest." 151 Here he has left an inference implicit which may be 
reconstructed by consulting later Kalam works. 152 He should have gone on to 



14t, K. al-Tamhid, p. 17. 
'"Above, p. 136. 

148 K. al-Tamhid, pp. 18-19. For various views on the origin of the term ma'na, see Wolfson, 
The Philosophy of the Kalam, pp. 147-167. 
""Cf. below, p. 180. 

150 AT. al-Tamhld, p. 22; cf. Maimonides, Guide, I, 73 (7). 
"'Above, p. 136. 

" 2 Cf. 'Abd al-Jabbar, Sharh al-Usul. (Cairo, 1965). pp. 93-94, 104. and 107; idem, K. al- 
Majmu", p. 53; Juwaynt, K. al-lrshad, p. 20; idem, K. al-Shdmil, pp. 186-187. The attribution of 
the Sharh al-Usfd to 'Abd al-Jabbar is not certain; see the editor's remarks in his introduction, pp. 27- 
28. 



138 



Kalam Proofs for Creation 



explain that whatever is subject to destruction is also subject to generation — the 
Aristotelian principle was accepted by the Kalam and is used by BaqillanI 
elsewhere 153 — and therefore the accident of motion, if subject to destruction, 
must be generated. The inference apparently was so familiar to BaqillanI that he 
simply takes it for granted. He makes no reference to it and passes at once to 
another matter, writing: If motion "were not destroyed with the advent of rest, 
both would be present together in the body, . . . something necessarily known 
to be impossible." At first glance the addition seems superfluous, 154 since when 
a body is at rest, motion obviously is not present; and once again we must consult 
later works to discover the import of the remark. BaqillanI is forestalling a 
possible objection to his statement that when the accident of rest is present, the 
accident of motion ceases to exist. The objection would be that the accident of 
motion may not completely cease to exist but may rather revert to a state of 
latency within the body, from which it can be elicited at a future time. In other 
words, the accidents of motion and rest are, perhaps, not generated after all, but 
are continually present in each given body, one of them always being in a state 
of latency, from which and to which it alternately emerges and returns. 155 Baqil- 
lanI forestalls the objection by pointing out, in effect, that the theory of latency 
is absurd since it transgresses the principle that contraries cannot be present in 
the same thing at the same time. 156 He is accordingly justified in affirming that 
when an instance of the accident of motion disappears, it must have been destroyed 
and hence must also have previously been generated. By an implied generaliza- 
tion, BaqillanI thereupon assumes that all instances of the accident of motion — 
not merely instances of motion that are seen to be succeeded by rest — and indeed 
all instances of all accidents must be generated. 

As a further step in his proof, BaqillanI undertakes to establish the proposition 
that bodies "do not precede generated things nor exist before them." He reasons 
that every body has its parts either joined, that is to say, in a state of composition, 
or separated. The reader is expected to understand that composition and separation 
are both accidents; 157 that by the earlier implied generalization all accidents arc 
generated; and therefore, since bodies are necessarily associated with either com- 
position or separation, they necessarily are associated with, and do not exist 
before, "generated things." The full contention, illustrated as here by the acci- 
dents of composition and separation, reappears in the same context in later Kalam 

'"Cf. Aristotle, De Caelo I, 12, 282b, 2; K. al-Tamhid, p. 20. 

I54 A similar statement appears in Philoponus' proof; cf. above, p. 92: "The same matter cannot 
admit several forms at once." 

l55 Thc theory of latency is reported in the name of Nazzam; cf. Khayyat, K. al-Inlisdr, §90; 
Wolfson, The Philosophy of the Kalam, pp. 498 ff. 

" 6 The same objection to the proof is dealt with by Baghdad!, K. Usui til-Din (Istanbul, 1928), 
p. 55; 'Abd al-Jabbar, Sharh al-Usul, p. 105; idem, K. al-Majmit ', p. 50 (only briefly); Juwaynl, K. 
al-lrshad p. 20; idem, K. al-Shamil, p. 190; Ghazali, al-lqtisiid, p. 28 (no solution is given). 

" 7 Cf. Pines, Alomenlehre, pp. 6-7. 



Kalam Proofs for Creat ion 



139 



versions of the proof. 158 In BaqillanI, at least, the introduction of the accidents 
of composition and separation is awkward. Since the accidents of motion and 
rest had been considered earlier in the proof, his presentation would have been 
tidier had he retained the original pair of examples and, having already shown 
that motion and consequently also rest are generated, contended now that all 
bodies necessarily arc associated with either the one or the other. Alternatively, 
he could have used the illustration of the accidents of combination and separation 
throughout with no change in the argument. Inasmuch as the use of two different 
pairs of accidents serves no function, it may be presumed to be due to features 
of the proof's tradition which arc no longer available to us. Among the early 
versions of the proof, Alfarabi's version too, as will be recalled, adduced the 
composition of bodies when establishing that bodies are always associated with 
accidents. 159 The appearance of composition in the proof from accidents might 
be explained with the aid of my suggestion that the proof grew out of the second 
of the auxiliary arguments whereby Philoponus supported his proof from the 
finite power of finite bodies. Side by side with that auxiliary argument Philoponus 
had offered an auxiliary argument from composition;"' 0 and in the course of 
transmission the reference to composition may have infiltrated from one argument 
to the other. 161 

BaqillanI, in sum, employs the standard reasoning whereby the Kalam estab- 
lishes the existence of accidents. He partially states and partially alludes to stand- 
ard reasoning for establishing the generation of accidents, the absurdity of the 
theory of latency, and the necessary association of bodies with "generated things." 
He rests his proof on the implied generalization from a single sort of accident — 
motion that is seen to be succeeded by rest — to all instances of all accidents. 
And he includes one element, the composition of bodies, for what are probably 
historical rather than intrinsic reasons. The proof must have come to him in a 
fixed, stylized form, around which a fund of theoretical discussion had already 
grown up. Either through superficiality or for the sake of brevity, he makes 
statements (hat are incomplete in themselves but take on meaning when under- 
stood as allusions to familiar discussions. 

A short time after BaqillanI, the Mu'tazilite 'Abd al-Jabbar (d. 1025/26) char- 
acterizes the proof we are examining as the most "dependable" one. He, inci- 
dentally, is the source of the report tracing the proof to the early Mu'tazilite Abu 
al-Hudhayl — who, 'Abd al-Jabbar adds, was followed in the use of the proof by 



l5B Cf. Baghdad!, K. Usui al-Din, pp. 49-50 (somewhat different); 'Abd al-Jabbar. Sharh al- 
Usitl, pp. 111-112; idem, K. al-Majmu', p. 50; Juwaynl, K. al-lrshad p. 24; idem, K. al-Shamil, 
pp. 204-205. 

" 9 Cf. above, p. 134. 

"'"Cf. above, p. 92. 

"''in Stoic physics, composition and separation arc forces for the production of new objects, and 
Kpicurus also had a doctrine of the mixture of atoms. 



140 



Kalam Proofs for Creation 



"the other sheikhs." 162 'Abd al-Jabbar's formulation or the proof is more syste- 
matized than Baqillani's, being explicitly "constructed upon four premises 
(da'dwa)." 163 Each of the four premises turns out to be a parallel of one of the 
steps in Baqillani's argument, but in BaqillanI neither steps nor premises were 
designated as such and the structure of the argument was not articulated sharply. 
The four premises enumerated by 'Abd al-Jabbar are (1) that "things" (ma'anin) 
are present in bodies, to wit, joining and separation, motion and rest; (2) that 
these are generated; (3) that body cannot be free (yanfakk) of them, nor does it 
precede (yataqaddam) them; 164 and (4) that what cannot be free of, or precede, 
what is generated is likewise (mithlaha) necessarily generated. 

The formalization of the four premises upon which the proof is constructed 
predates 'Abd al-Jabbar. A text entitled K. al-Majmu' fi al-Muhlt bi-l-Taklif, 
based on 'Abd al-Jabbar's thought and edited sometime after his death, 165 reports 
that the fully developed form of the proof from accidents had been known to the 
Mu'tazilite Abu Hashim (d. 933), the son of al-Jubba'I. According to the Majmu', 
Abu Hashim had insisted that the generation of body, and hence the creation of 
the world, can be proved only through the "four principles (usul)"; and he had 
declared that anyone denying the existence of accidents will be unable to accom- 
plish the proof. 166 For its part, the Majmu' — which, as just mentioned, was not 
written by 'Abd al-Jabbar but is based on his thought — does not merely use the 
four "principles" or "premises" to prove creation. 167 It offers elaborate argu- 
mentation to support the premises and to remove a variety of possible objec- 
tions; 168 and it accompanies the proof with a running methodological discussion, 
by both 'Abd al-Jabbar and his posthumous editor, dealing with the questions 
whether the present proof is in truth the best way of establishing the creation of 
the world and whether the best procedure for proving the existence of God is by 
proving creation first. 169 By the time of the Majmu ' a theological adept apparently 
could take the proof for granted, and could busy himself with methodological 
issues and the dialectic overgrowth that had enveloped the proof. 

The proof of creation from accidents is employed by the Ash'arites BaghdadI 
(d. 1037) 170 and Juwaynl (d. 1085). 171 Juwaynl establishes the creation specifi- 
cally of "substances," that is, of atoms rather than bodies; his proof therefore 
concludes with the creation of the world ex nihilo and not merely the creation of 

,62 Snarh al-Usul, p. 95. 

163 Ibid. On p. 104, 'Abd al-Jabbar refutes the theory of latency as part of a proof of the generation 
of accidents. 

'"Regarding these terms in the early version of the proof, cf. above, p. 136. 

l65 Cf. the editor's French introduction, p. 8. 

'"AT. al-Majmu' , pp. 30. 63. 

167 The proof appears in K. al-Majmu', pp. 30 ff. 

"■"Ibid., pp. 32-58. 

169 Ibid., pp. 28-31. 

"°K. Usui al-Din, II. 

"'K. al-lrshad, pp. 17-27; K. al-Shamil, pp. 166 ff.; Textes apologitiques {Luma'), pp. 120 ff. 



Kalam Proofs for C real ion 



141 



the world in its present form. Like 'Abd ai-Jabbfir, Juwaynl sets down four 
principles, but he makes a significant change, to be discussed in the next section. 
The proof is employed as well by MawardI (d. 1058) 172 and by Bazdawl (d. 
1099), a follower of Maturidl. 173 Juwaynl's version reappears in Ghazali's Kalam 
writings 174 and in Shahrastanl. 175 The proof from accidents is canonized, as it 
were, in the creed of Nasafi (d. 1 142/43). 176 It is employed by the Isma'ili Nasir- 
i-Khosraw 177 and is cited by Abu al-Barakat 178 and Ibn Tufayl. 179 Averroes, an 
outspoken opponent of the Kalam, 180 and also TusI (d. 1273), 181 Amid! 182 and 
Iji 183 describe it as the Kalam proof for creation par excellence. 

The proof enjoyed a good deal of popularity among Jewish followers of the 
Kalam. The Karaite Joseph al-Basir (tenth or eleventh century) explicitly con- 
structs it on four "premises," 184 and his student Jeshua b. Judah employs it as 
one of several methods for proving creation. 185 Among the Rabbanites, it appears 
in Joseph Ibn Saddiq (d. 1149), 186 in a work attributed to Abraham Ibn Ezra 
(1092-1 167), 187 in Judah Hallevi, 188 and in Joseph Ibn Aqnin (Joseph b. Yahya) 
(d. 1226), where it is described as the "current" (or perhaps "preferable" [mer- 
usa\) method of demonstrating creation. 189 The formulations in the works of the 
two Karaite writers just mentioned and in the work attributed to Ibn Ezra disclose 
a strong resemblance to 'Abd al-Jabbar's formulation. Finally, Maimonides, like 
Averroes an outspoken opponent of the Kalam, includes Juwaynl's version in a 
list of seven Kalam proofs of creation. 190 



,72 M5wardi, A' lam al-Nubdwa (Cairo, 1971), pp. 13-14. 
'"tf. Usui al-Din, ed. II. Linss (Cairo, 1963), p. 15. 

"'al-Iqtisad. pp. 24-32; al-Risala al-Qudsiya, pp. 16-17, 34-35. Cf. Tahafut al-Falasifa, 
X, SI. 

n, K. Nihaya al-lqdam, p. 11. 

" 6 E. Elder, A Commentary on lite Creed of Islam (New York, 1950), pp. 28-29 (Nasafi) and 
p. 33 (Taftazanl's commentary). Cf. A. Wensinck, "Les Preuvcs de l'Existence de Dieu," Mededee- 
lingen der Koninklijke Akademie van Wetenschappen le Amsterdam, LXXXI, Series A, No. 2 (1936), 3. 

,77 Cf. Pines, Atomenlehre, p. 37, n. 2. 

178 AT. al-Mu' labor. III, p. 31. 

n9 Hayy ben Yaqdhan, Arabic text, p. 81; French translation, p. 62. 

180 K. al-Kashf, pp. 31-32. Averroes enumerates only three premises, by not counting the existence 
of accidents as a separate premise. 

""Glosses to Razi's Muhassal p. 89 K2 Ghaya al-Maram pp. 261-262. 

m Mawaqif. VII, p. 222. " 

l84 P. Frankl, Ein Mu' lazilitischer Kalam (Vienna, 1872), pp. 20, 53. 

,85 M. Schreiner, Studien ueber Jeschu'a benJehuda (Berlin, 1900), pp. 29-33. On p. 31, n. 2, 
Schrcincr adds further references to the use of the proof. 

>u ha-'Olam ha-Qatan, cd. S. Horovitz (Breslau, 1903), pp. 48-49. 

181 Kerem Hemed, IV (1839), 2-3. Cf. M. Schrcincr, Der Kalam in der juedischen Literatur 
(Berlin, 1895), pp. 37-40. 
lu Kuzari, V, 18 (2). 

,s9 Trealise as to Necessary Existence, cd. and trans. J. Magnes (Berlin, 1904), Hebrew section, 
pp. 17-19; English section, pp. 39-42. 
,90 Guide, I, 74 (4). 



( 



142 Kalam Proofs for Creation 

In at least two writers, Shahrastanl and Aaron ben Elijah, the Aristotelian 
matter and form appear in place of body and accident, with the result that the 
argument approaches or reapproaches the argument of Philoponus from which 1 
have suggested it derives. 

Shahrastani knew of a peculiar tradition according to which Aristotle had 
espoused the creation of the world. 191 And he attributes to Aristotle an argument 
that appears simply to be the Kalam proof we have been examining, with the 
dichotomy of matter and form substituted for the dichotomy of body and accident. 
Aristotle, Shahrastanl would have us believe, reasoned: "Forms ... are not from 
one another, but are necessarily after one another, so that they succeed one 
another over matter, and plainly are destroyed and pass away. Whatever passes 
away must have a beginning; for passing away is an end, and it is one of two 
terminuses, to which a counterpart must correspond. Now it has been established 
that the [forms] are something generated out of nothing 192 and that their subject 
is essentially disposed to receive them and serve as their subject. They have a 
beginning and an end; this shows that their subject has a beginning and an 
end. . . ." I93 

Another formulation in terms of matter and form is given by the relatively late 
Jewish Karaite writer Aaron ben Elijah (d. 1369), who undoubtedly wished to 
free the proof from the burden of Kalam physics and render it more respectable. 194 
In the first of four proofs for creation, Aaron begins by establishing the existence 
of matter and form. To that end he offers a peculiar adaptation of the reasoning 
i' used by the Kalam thinkers for establishing the existence of accidents. Kalam 

( reasoning, applied to the accidents of composition and separation, 195 would run: 

Bodies cannot be in a state of combination or separation by virtue of themselves, 
'{ since if such were the case, they would either always be in a state of combination 

j{ or always in a state of separation, whichever their own nature necessitates; con- 

| sequently, both combination and separation must be "something" (ma'nd) added 

j( to bodies, that is to say, they are accidents. Aaron alters the meaning and role of 

\( combination and separation, reasoning: "Combination and disjunction are con- 

| traries, and matter [read: body] receives both. What receives conjunction cannot 

be the cause of disjunction; rather we must conclude that combination is due to 
■(■ 'something' ('inyan = ma' no) and disjunction is due to 'something' else. Body 

| '"AC. al-Milal wa-l-Nihal, p. 326. Shahrastanl realizes lhat the tradition is out of harmony with 

/ the position generally attributed to Aristotle, and he too sometimes attributes the doctrine of eternity 

to Aristotle; cf. ibid., pp. 320-321, 338, 340. 
:( l92 Cf. Philoponus' statements, above, p. 92. 

! , m K. al-Milat wa-l-Nihal, pp. 326-327. 

' l94 0nc of Maimonides' objections to the proof is that it rests on the theory of atomism, which is 

'.j not accepted by those who believe in eternity (cf. Guide, I, 74 (4)), and Aaron ben Elijah was 

completely familiar with the Guide. 
( l95 BaqillanT applied this reasoning to the accident of motion, and 'Abd al-Jabbar applied it to the 

accidents of composition and separation as well. Cf. above, nn. 148, 163. 



Kalam Proofs for Creation 



143 



thus has Iwo 'things' Cinyanim), matter and form, disjunction coming from the 
former, and conjunction from the latter." Aaron hereupon proceeds in the 
familiar manner, contending that forms arc generated, that matter is never free 
of forms, and that matter therefore is likewise generated. To justify his conclusion 
he cites those whom he calls "our scholars," who "have stated that if anything 
cannot be free of (yit 'arch) what is generated, it likewise (kamohu) is generated." 197 

4. Juwayni's version of the proof from accidents 

At the the heart of the proof from accidents lies the contention that the subject 
in which generated accidents are ever present must also be generated. The con- 
tention is by no means self-evident and unless it is justified in some fashion the 
conclusion of the proof remains unfounded. A development in the approach to 
the key contention can be discerned among the philosophers who employ the 
proof from accidents. 

Frequently the bald assertion was made that whatever is unavoidably associated 
with what is generated is itself generated, and the argument was left at that. 
Saadia's statement of the argument, for example, concludes: "What cannot avoid 
what is generated is known to be of the same character." 198 Alfarabi deliberately 
draws out the argument, with which he himself has no sympathy, into seven steps; 
but they all merely amount to the assertion that whatever is necessarily joined to 
what is generated is itself generated.' 99 BaqillanI seems to add something new 
when he writes that whatever "docs not precede what is generated is likewise 
generated" because it must either come into existence "together" with, or "after" 
its generated concomitant; 200 but the statement is cither a tautology or, as will 
be seen, an equivocation. 'Abd al-Jabbar declares: "Since body is not free of the 
aforementioned generated things and docs not precede them, its lot (hazz) in 
existence must be like their lot," and he illustrates the assertion by the fact that 
twins are always of the same age. 201 Baghdad! states: "Inasmuch as bodies do 
not precede generated accidents, their own generation is entailed; for what does 
not precede generated things is generated, just as what does not precede any 
single generated thing is generated." 202 

The weakness in all these statements is, as Averrocs points out, the danger of 
equivocation in such expressions as whatever is joined to what is generated and 
whatever does not precede what is generated; for the words what is generated 
can mean cither a single generated thing or else generated things. Obviously, if 
any body is necessarily joined to a particular accident that is generated, the body 

''"'That is to say, form brings things together into a single class, and matter is the principle of 
individuation. 

m 'Es Hayyim, chapter 10, p. 28. 

198 Above, p. 104. 199 Cf. above, pp. 134-135. 

2 ""Above, p. 136. 

2a[ Sharh al-Usul. pp. 1 13-114; cf. A'. al-MajmiV , p. 58. 
202 A.'. Usulal-Din, pp. 59-60. 



144 



Kalam Proofs for Creation 



does not exist before the accident, exists together with it, and also is generated. 
It is far from obvious that a body must be generated if necessarily joined to one 
generated accident or another, but not to a particular, given accident; and precisely 
the latter proposition is at issue. 203 Why — as Ibn Suwar expressed his objection 
to the proof 204 — can the body of the universe not be joined from all eternity to 
an infinite series of generated accidents? 

Baghdad! may have been cognizant of the weakness in the proof. In his con- 
clusion he writes that "what does not precede generated things is generated, just 
as what does not precede any single generated thing is generated." 205 The state- 
ment seems to be an assurance that the proof docs not suffer from the equivocation 
to be brought out later by Averroes. Baghdad! fails, however, to explain the basis 
of his assurance. In ' Abd al-Jabbar's Sharh al-Usul and in the Majmu ' , an objec- 
tion identical with lbn Suwar's is explicitly taken up after the proof is completed. 
In the Sharh al-Usul, the objection reads: "Why deny that although body is not 
free of generated things. . . , one generated thing was generated in it, prior 
thereto another, prior thereto yet another, ad infinitum!" 206 In the Majmu ' the 
same objection is introduced by the words: "In order to perfect the proof of the 
creation of bodies, there is no avoiding . . . showing that generated things have 
a first [term]." 207 Solutions are given in both works, and finally the Majmu' 
comes to the solution that was to change the character of the proof. 'Abd al- 
Jabbar, according to the Majmii ', "demonstrated the finiteness of these generated 
things by the presence in them of increase and diminution"; 208 that is to say, he 
contended that the accidents passing over the body of an eternal universe could 
not form an infinite series, because the argument from "increase and diminution" 
shows an infinite series to be impossible. 'Abd al-Jabbar thus answered the critical 
objection to the present proof by adducing what had originally been a separate 
proof, namely, one of the proofs for creation from the impossibility of an infinite 
number. 209 

It is Juwayn! who fully and explicitly recognizes that what is at issue here is 
not merely one of several difficulties to be raised and answered dialcctically, but 
rather the nerve of the entire demonstration. As in the tradition summed up in 
'Abd al-Jabbar, 210 Juwayn! bases the proof from accidents on "four principles." 
First he establishes the existence of accidents, then shows them to be generated, 
and thirdly shows that no substances — that is, no atom — can be free of them. 



K. al-Kashf, p. 35. A similar critique is made by Abu al-Barakat, K. al-Mu'tabar, III, p. 31. 
204 See above, n. 139, Arabic text, p. 245; French translation, §5. 
205 Above, n. 202. 
2M Sharh al-Usul, p. 1 14. 
207 K. ai-Majmu\ p. 59. 
20R Ibid., p. 61. 

2<w Baqillani may be alluding to the same type of response to the objection, K. al-Tamhid, p. 49, 
taken together with p. 25. 
2l0 Above, p. 140. 



Kalam Proofs for Creation 



145 



But he no longer feels he can assign the status of a principle to the proposition 
that what is unavoidably associated with generated accidents is itself generated. 2 " 
As his fourth premise, he substitutes "the impossibility of generated things with- 
out a first |term]," 212 a doctrine that, he goes so far as to aver, is the touchstone 
separating the believer from the nonbeliever who maintains the eternity of the 
world. 213 Juwayn! supports his fourth premise as follows: "It is a principle of 
the heretics that prior to the present revolution (of the sphere] infinite revolutions 
have been completed. But for something infinite to pass away by having one unit 
succeed another is impossible. The fact that the revolution[s] [of the sphere] prior 
to the present rcvolution(s) have passed away, . . . demonstrates their finite- 
ness." 214 This is simply the familiar and much repeated argument that the world 
must have a beginning since an infinite series of past events cannot conceivably 
have been traversed. 215 Juwaynl's complete demonstration accordingly runs: The 
universe consists of atoms and accidents. Accidents arc generated; and atoms arc 
never free of them. But an infinite series of generated accidents is impossible 
since an infinite cannot be traversed, and hence there must be a first, generated 
term for every series of accidents. Inasmuch as each atom and the universe as a 
whole arc inextricably associated with finite series of accidents, each atom and 
the universe itself can have existed for only a finite time, and must have come 
into existence from nothing. Juwaynl's version rests, as Averroes later was to 
demand the proof must, 216 on the association of the universe with a single gen- 
erated thing, that single generated thing being the series of generated accidents — 
or, to be more precise, any single series of generated accidents. 217 — in the universe. 

To recapitulate: The key contention in the Kalam proof from accidents cries 
out for some justification. The lack was eventually felt, and was filled by adducing 
one or another of the arguments — it makes no difference which — from the impos- 
sibility of an infinite number. The classic Kalam proof for creation, which may 
have derived from Philoponus, was thereby combined with the proofs for creation 
from the impossibility of an infinite number which undoubtedly did derive from 
him. Perhaps it would be more accurate — seeing that the proof from accidents 
failed by itself to demonstrate anything — to say that the entire burden of dem- 
onstration fell upon the proofs from the impossibility of an infinite number. The 
proof from accidents regressed into a mere prelude to those proofs. 

The version of the proof from accidents appearing in Ghazali, ShahrastanI, 
Maimonides, and Tus!, 218 is Juwaynl's version. Averroes, in a critiaue of the 
Kalam, first records the earlier version of the proof, discussed in tne previous 

2n Hc does know the proposition; cf. A". al-Shamil, pp. 220-221. 

2>2 K. al-Irshad, p. 18; K. al-Shamil, p. 215; Texles apologeliques (Luma'), p. 120. 

2l1 Af. al-lrshad, p. 25. 

2l4 Ibid., p. 26: cf. K. al-Shamil, p. 215; Texles apologeliques (Luma 1 ), p. 126. 
2 "Above, pp. 119-120. ^ Above, p. 144. 

2l7 Such as the movements of the heavens. 
2l8 Above, nn. 174, 175, 181, 190. 



146 



Kalam Proofs for Creation 



section. Then he adds: "After the later Kalam thinkers realized the weakness of 
the proposition [that what is not free of generated things is itself generated], they 
tried to strengthen it by proving, as they supposed, that an infinite number of 
accidents cannot pass successively over a single subject." 219 The later Kalam 
thinkers, Averroes continues, based the impossibility of an infinite series of acci- 
dents on the contention that if any event had to be preceded by an infinite number 
of events, it could never come about; in other words, they employed the argument 
of proofs from the impossibility of an infinite number. Such, we have seen, is 
precisely the procedure followed by 'Abd al-Jabbar and more formally by Juwaynl. 
After Averroes records the version of the proof from accidents attributed by him 
to the "later Kalam philosophers" he sets forth his own refutation. 220 

The manner in which Shahrastanl and Averroes refer to what 1 have here called 
Juwaynl's version of the proof from accidents, indicates that they did not consider 
him to be its author. 221 Juwaynl, though, docs claim credit for the new 'version 222 
and is the earliest writer in whose works 1 have been able to find it fully articulated. 

5. Proofs from composition 

Philoponus, as was seen in the previous chapter, offered two arguments from 
composition in support of his overall proof for creation from the finite power of 
the physical universe. Both arguments take their departure from the composition 
of matter and form in the heavens. In one of them, the reasoning was that what 
is composite is not self-sufficient, hence not infinitely powerful, hence destruc- 
tible and generated. 223 The reasoning in the other was that what is composite is 
subject to decomposition, and what is subject to decomposition is not infinitely 
powerful, the final inference of creation from finitencss of power not being stated. 224 
The line of thought of the second of the two arguments goes back to the Phaedo, 
where being composite is held to imply that a thing is subject to decomposition. 225 

The entire enterprise of citing composition in the universe as evidence of 
creation may well have been prompted by a passage in Aristotle. An argument 
is found in Aristotle to the effect that anything composite has the possibility of 
existing in a noncomposite state, and that the possibility of existing in a noneom- 
posite state, like every possibility, must have been realized at some moment in 
past time. Aristotle inferred herefrom that "the eternal things" cannot be "com- 
posed of elements" and consequently do not contain matter and form. 226 Philo- 
ponus would plainly proceed to the further conclusion that since eternal things 

"*K. al-Kashf, p. 36. 
220 Ibid., pp. 36-37. 

22l Thcy both contrast the present proof with another that they do attribute to Juwaynl. References 
above, nn. 175, 220. 
222 K. al-Shamii p. 218. 
223 Above, p. 92. 
224 Above, pp. 92-93. 
22, Phaedo, 78C; above, p. 92. 
™ Metaphysics XII, 2, 1088b, 14-28. 



Kalam Proofs for Creation 



147 



are not composite whereas the heavens are, the heavens cannot be eternal. A 
different conclusion, more in the spirit of Aristotle's philosophy, might, however, 
be drawn with equal cogency. The proper conclusion to be drawn from Aristotle's 
words is, according to Averroes, not that the heavens arc generated but that they 
arc free of the composition of matter and form: If eternal beings are not composite 
and the heavens arc known with certainty to be eternal, then — as Averroes spells 
out Aristotle's intent — the heavens arc not composite. 227 

An argument is also to be found in Proclus which concludes that since the 
physical universe is composite, it must be "generated." But what Proclus means 
thereby is that the universe is generated eternally, that is to say, eternally depen- 
dent on a cause for its existence. He reasons: Anything consisting in the joining 
together of parts must have a cause responsible for the parts' being joined; never- 
theless, the joining together of the parts and the composite product may be eternal, 
as is the case, so Proclus understands, with respect to the heavens. 228 In arguing 
from the composition of the physical universe to the generation of the physical 
universe, Proclus is, in other words, advancing a proof of the existence of God 
and not a proof of creation. 

In medieval Arabic philosophy a number of arguments from composition are 
in evidence. Some are arguments for the creation of the world, and others, for 
the existence of God, without reference to creation. The character of the com- 
position from which cither creation or the existence of God is derived varies, and 
at least three strains can be distinguished, (a) In certain instances the general fact 
of composition in the universe is adduced with no attention given to the specific 
character of the composition. Here an additional bifurcation can be discerned. 
Sometimes the inference of creation or the existence of God is put forward without 
explanation and apparently is regarded as self-explanatory or self-evident; such 
arguments may stem from Philoponus or, conceivably, even directly from Plato, 
Aristotle, or Proclus. Sometimes, by contrast, grounds are given for the inference 
of creation or the existence of God from composition, (b) In other instances, the 
combination specifically of contrary qualities in the universe is focused on, the 
contention being that the joining together of contrary qualities could only be 
effected by an overriding external force, (c) In still other instances, the focus is 
on the purposefulness of the composition in the universe, and the arguments arc 
of a teleological character. None of the arguments in any of the categories is 
particularly subtle or profound. 



Middle Commentary on Metaphysics, Casanatense Library, Hebrew MS. 3083, XII, p. 140 
(14l)a; Tahafut al-Tahiifut , IV, pp. 280-281. 

^Commentary on Timaeus, cd. L\ Dichl, Vol. I (Leipzig, 1903). pp. 290, 297. French translation, 
with pagination of the Greek indicated: Commentaire sur le Timee, trans. A. Fcstugierc, II (Paris, 
1967). Elements of Theology, cd. and trans. E. Dodds (Oxford, 1936), §47; Liber Dc Causis, ed. 
and German trans. O. Bardenhewcr (Freiburg, 1882), §27. 



( 
( 

I 

148 Kalam Proofs for Creation 

( (a) In the previous chapter, Saadia and Kindi were seen to offer proofs for 

creation from the composition in the universe. Saadia's proof did not explain how 

' composition implies creation, 229 and Kindi's was discovered to be convoluted 

and circular. 230 Because of the context in which Saadia presented his proof and 
because of Kindi's terminology and the context in which he presented his, 1 

< suggested that both were outgrowths of Philoponus' arguments from composi- 
, tion. 231 A proof of creation from composition is also to be found in the 'Uyun 

al-Masd'il, a work attributed — incorrectly — to Alfarabi. The following syllog- 
( ism appears there: "The world is composite; everything composite is generated; 

consequently ... the world is generated." 232 As in Saadia and Kindi, only cre- 
ation, not creation ex nihilo, is established by the argument; and, as in Saadia, 

< the inference of generation, or creation, from composition is in no way justified. 
, Another argument from composition in which the inference remains unexplained 
' is advanced by the Arabic Christian writer Theodore Abu Qurra (ca. 740-820); 
I now, however, the existence of God is being proved, although with an intimation 

that creation too might be proved by the argument. Abu Qurra lays down the 

principle that "whenever something is composite, its parts precede it naturally 
{ and usually . . . also temporally." 233 To buttress the principle, he refers to the 

empirical truth that the existence of building materials "precedes" the construc- 
^ tion of a house; but he nowise essays to demonstrate the principle philosophically. 

( His conclusion is that a composite universe must depend for its existence upon 

an external agent who binds the components together. 234 Similar arguments for 
' both the existence of God and the creation of the world appear among the early 

I Scholastics. 235 

Attempts were sometimes made to explain why the existence of God and the 
( creation of the world do follow from the composition of the universe. ShahrastanI 

( cites an argument in the name of Ash'ari which utilizes Kalam concepts for the 

purpose. Atoms, the reasoning runs, continually pass from a state of being joined 
( with other atoms to a state of being disjoined from them. Their state of being 

/ joined or disjoined cannot flow from their own nature; for if the nature of a given 

' ""Above, p. 103. 

( 230 Above, pp. 111-112. 

"'Above, pp. 102-103, 115. 
( 232 ' Uyun al-Masa' il, §2, in Alfarabi's philosophische Abhandlungen, cd. F. Dietcrici (Leiden, 

1890). German translation: Alfarabi's philosophische Abhandlungen aus dem Arabischen uebcrsetit, 
1 trans. F. Dietcrici (Leiden, 1892). 

233 For the different senses in which one thing can be said to precede another, see Aristotle, 
' Categories 12. 

| 234 Arabic text in al-Mashriq, XV (1912), 762; German translation: Des Theodor Abu Kurra Traklat 

Ueber den Schoepfer, trans. G. Graf (Muenster, 1913), pp. 16-17. 
/ 235 Cf. G. Grunwald, Geschichte der Gottesbeweise im Miltelaller (Muenster, 1907), pp. 57-60 

(arguments for the existence of God where the proposition that every composite has a cause is treated 
| as self-explanatory); pp. 62 - 64 (where that proposition is explained by Alan of Lille); pp. 68-69 

(an argument for creation where the inference is again treated as self-explanatory). 

( 



Kalam Proofs for Creation 



149 



atom determined that it should be joined, it would always be so, and if its nature 
determined that it should be disjoined, it would always be disjoined. 236 The 
passage of atoms from one state to another must therefore depend on an external 
agent who is responsible for assigning either the one or the other state to each 
given atom. Thus far, Ash'ari and ShahrastanI have provided a crude proof of 
the existence of God. To carry the argumentation forward and develop it into a 
proof of creation, ShahrastanI adds a principle met earlier, the principle affirming 
that "what docs not precede what is generated is likewise generated." 237 The 
conclusion is not rendered explicit by ShahrastanI, but is clear: Since atoms do 
not precede the states of conjunction and disjunction, and since every state of 
conjunction or disjunction is generated, all atoms and therefore the universe as 
a whole must be generated — generated, be it noted, ex nihilo. 21 * The argument, 
which is cited by ShahrastanI in the name of Ash'ari, is recorded by Maimonides 
in his enumeration of the Kalam proofs of creation. There is a difference, how- 
ever, in that Maimonides docs not venture to explore the argumentation beyond 
the stage where an external agent responsible for the conjunction and disjunction 
of atoms is inferred. Maimonides docs mention, with a hint of disdain, that the 
Kalam thinkers thought the argument could establish creation, as well as the 
existence of God. But the step in which creation was reached was apparently 
judged by him to be too weak to waste words on. 239 

Once Aviccnna appeared on the scene, a new route was made available for 
completing arguments from composition. Aviccnna, as will be seen in a later 
chapter, devoted considerable effort to the analysis of the concept possibly exis- 
tent. And in the course of analyzing the concept, he established that every com- 
pound is possibly existent and that everything possibly existent depends upon 
something outside itself for its existence. 240 Those propositions, taken together 
with the empirical fact that the world is compound, furnished Fakhr al-DIn al- 
RazI with the materials for a proof of the existence of God. The physical universe, 
Rftzl argues, is compound, everything compound is possibly existent, and every- 
thing possibly existent has a cause of its existence; consequently, the physical 
universe has a cause of its existence. 241 Whereas Razi's reasoning concludes with 
the existence of God, Ijl presented similar reasoning as a proof of both the 



236 The reasoning would have been tighter, had ShahrastanI written: "If the nature of a given atom 
determined that it should be joined to a particular other atom. ..." In Kalam physics, all atoms in 
the world are in combination. 

217 Above, p. 136. . 

" K K. Nihaya al-lqdam. p. 1 1; Wolfson, The Philosophy of the Kalam, p. 386. 

219 GwiWe, I, 74 (3). The conclusion that atoms arc created depends on the questionable principle 
that "what docs not precede what is generated is likewise generated " Cf. above, p. 143. Maimonides 
himself offers a proof for creation from the composition in the universe. Cf. Guide. II, 22; below, 
p. 208. 

240 Below, p. 296. 

24 'Mului.ual, p. 107. 



( 
{ 

( 

150 Kalam Proofs for Creation 

existence of God and creation. The physical universe, Ijl contends, is compound; 
everything compound is possibly existent; what is possibly existent has a cause 
( that brought it into existence; but whatever is brought into existence is preceded 

by nonexistence and hence is generated. 242 

(b) A second strain of argumentation does not concern itself with composition 
in general but rather with the composition specifically of contrary qualities. The 
notion that God or Nature reconciles the contrary forces in the universe goes 
back at least to the Pseudo-Aristotelian De Mundo. 243 A proof of the existence 
( of God using the notion goes back at least to Athanasius (fourth century). Athan- 

asius asks rhetorically how the existence of God can be denied by anyone who 
"discovers fire mixed with the cold, and the dry mixed with the wet, yet not 
( opposing one another," that is to say, by anyone who beholds the blending together 

of the elements in nature despite their contrary qualities. The elements, with their 
opposing qualities, could, Athanasius concludes, only be bound together by an 
overriding external agent, in other words, by a deity. 244 Similar thinking appears 
in John of Damascus, Theodore Abu Qurra, and Job of Edessa, each of whom 
could have served as a bridge to the Islamic world. John of Damascus offers a 
( proof, again of the existence of God, not creation, which reads: "How could such 

contrary natures as fire and water, earth and air, combine with one another to 
form one world and remain undissolved unless there were some all-powerful 
( force to bring them together and always keep them so?" 245 Abu Qurra was seen 

earlier to contend that an external agent must be posited who binds together the 
( components making up the universe. 246 Immediately after having made the gen- 

eral point, Abu Qurra turns his attention to the contrary characters of the four 
elements. He observes that earth and water naturally descend whereas air and 
( fire ascend, that, further, contrary elements naturally destroy one another; and 

still the elements are constrained against their contrary natures into combination. 
The conclusion he draws now is stronger than the conclusion he drew earlier: 
Not only does composition in the universe indicate a cause; it indicates a cause 
able to "constrain" and "bind" together contrary qualities "against their nature," 
hence a cause possessed of "immeasurable . . . and indescribable power." 247 Job 
( of Edessa, who wrote in Syriac, likewise contends that the contrary qualities — 

r "heat," "cold," "wetness," "dryness" — cannot be imagined to combine by "them- 

selves, because if left to themselves they would not have been induced to do 



( 



( 242 Mawaqif, VII, p. 227. Ij! writes that the argument comes from Razi. For other arguments to 

the effect that what is possibly existent must be created, sec below, pp. 191, 387. 

2ii De Mundo, §5. Cf. W. Jaeger, Nemesius (Berlin, 1914), p. 112, n. 2; K. Gronau, Poseidonius 
und die juedisch-christliche Genesisexegese (Leipzig, 1914), p. 143, n. 2. 

lu Contra Gentes, cd. R. Thomson (Oxford, 1971), §36. Cf. Bocthius, Consolation, III, prose 

12. 



I 



245 



De Fide Orthodoxa, I, 3. 



( 24fi Above, p. 148. 



247 Sec above, n. 234, Arabic text, pp. 762-763; German translation, pp. 17-18. 



Kalam Proofs for Creation 



151 



anything opposed to their nature. There must consequently exist a being beyond 
the elements" which combines them, "and that being is God." 248 

Kalam writers cite the combination of contrary qualities in the universe as 
evidence both of the creation of the world and the existence of God. The Mu'tazilitc 
Nazzam is reported to have proved creation thus: "I find heat and cold joined in 
a single body despite their contrariety and mutual divergence; and 1 understand 
that they cannot be joined by virtue of themselves . . . that the agent who joined 
them is the agent who created them joined, and who constrained them. . . ," 24 '' 
Again: "Heat, 1 find, is contrary to cold, and contraries cannot be joined through 
themselves in a single place. Having found that they arc joined, I know they have 
something that has joined and constrained them against their own character. What 
undergoes constraint ... is weak. And its weakness and the fact that the con- 
straining agent exercises effective control upon it are a proof of its having been 
generated and that something has generated and created it." 250 

MaturldT cites the composition in the universe to prove both the existence of 
God and the creation of the world. When treating of the existence of God, he 
notes that the world contains contrary qualities, and he continues: "Any object 
in which contrary . . . and divergent natures arc joined . . . cannot be joined by 
virtue of itself; consequently, it has a joining agent." 251 When treating of the 
issue of creation, he writes: "Perceivable objects [including the world as a whole] 
are unavoidably subject to the joining of differing and contrary natures. Those 
natures are characterized by mutual repulsion. . . . Consequently, their joining 
must be due to something other than themselves; and that establishes their 
generation." 252 

(t) In still other instances, attention is focused on the purposefulncss of the 
composition in the universe, be it the purposefulncss of composition in general 
or the purposefulncss of the composition of contrary qualities. 

Athanasius was just seen to infer the existence of God from the fact that 
contrary qualities are combined in the universe. 25 " 1 But as Athanasius proceeds 
we find that the joining of contraries implies for him not merely a "superior being 
and master" who forces them to obey as "slaves obey their lord." 254 The com- 
bination of contraries forms an "order" and "harmony," hence implies a "ruler," 
"director," and "king," who is responsible for the order and harmony. 255 The 
same configuration recurs in John of Damascus. John of Damascus, as was seen, 
infers the existence of God from the combination of "contrary natures." But he 



24H !iook of Treasures. I, iv. p. 15 (cited by A. Wry, Al-Kindi's Metaphysics | Albany, I974|. p. 27). 

M "Khayyat, K. al-inlisar, §26. 

""Ibid. 

"'A'. al-Tawltid. p. 18. 
" 2 lbid., p. 12. 
"■'Above, p. 150. 
""Contra Gentes. §37. 
"'Ibid.. §38. 



152 



Kalam Proofs for Creation 



prefaces his proof with the heading: "The very maintenance, preservation, and 
government of creation teaches us there is a God." And he incorporates into his 
reasoning the observation that the four elements are "arranged" in "unceasing 
and unhindered courses." 256 His argument from the combination of contraries 
has, accordingly, itself entered into combination: It is combined with, or but- 
tressed by, teleological considerations; and Aquinas was indeed to read it as a 
proof from design. 257 Scholastic literature exhibits additional instances of argu- 
ments for the existence of God where the composition of contraries is cited as 
evidence of design. 258 

The theme that the combination of contraries implies design appears in the 
writings of the Ikhwan al-Safa'. The Ikhwan recommend the study of plant life 
in all its vast variety on the grounds that "a well-made product indicates a wise 
maker." The accompanying explanation is that "the four elements, with their 
opposing powers and mutually antagonistic natures, could not have been joined 
and combined [in the plant realm] . . . except through the design of a wise 
maker"; therefore, the study of plants leads to knowledge of God. 259 

The most complex web is woven by the Jewish writer Bahya ibn Paquda. Bahya 
offers a proof of creation based on the principle that "everything combined is 
generated," since its parts are "naturally" and "temporarily" prior to the com- 
posite. 260 As supporting evidence, he cites three types of composition in the 
universe. He first notes the construction of the universe from different parts, 
which are "combined and composed" to produce a fully furnished dwelling place 
for man, the "householder." Here the theme clearly is teleological. Then he 
observes that plants and animals are composed of "the four elements, . . . which 
differ and exclude one another, so that there would be no way for man to combine 
and order them. ... Yet their combination ... is firm and stable. . . . They 
cannot be mixed by virtue of themselves . . . [and consequently] what does 
combine them is different from them." Here we have the consideration that con- 
trary qualities can be reconciled only through a constraining agent; and as in 
Athanasius and John of Damascus, the thought is associated with teleological 
considerations. As a third piece of evidence for composition in the universe Bahya 
observes that the elements themselves are "combined from matter and form, that 
is to say, from substance [i.e., atom] and accident; their matter is prime mat- 
ter, . . . whereas their form is the first general form which is the root of every 
substantial and accidental form." 261 The "first general form," the form common 



256 De Fide Orlhodoxa. I, 3. 

151 Summa contra Gentiles, I, 13. 

258 C. Baeumker, Wilelo (Muenster, 1908), p. 318. 

159 Rasa'it, Physics, vii (Beirut, 1957), Vol. II, p. 152. German translation: F. Dietcrici, Die 
Naturanschauung und Naturphilosophie der Araber im X Jahrhundert (Leipzig, 1 876), p. 163. 
260 Cf. Abu Qurra, above, p. 148. 

2< "at-Hidaya (Hobot ha-Lebabot) , cd. A. Yahuda (Leiden, 1912), I, 5 and 6. 



Kalam Proofs for Creation 



153 



to all elements, is nothing other than corporeal form, posited by certain com- 
mentators as an intermediate stage between Aristotle's prime matter and the 
proper form of each element. Corporeal form was construed by the Arabic Aris- 
totelians as identical with the tridimensionality of all physical objects, or as the 
medium for the presence of tridimensionality in prime matter. 262 In elucidating 
the composition of the elements, Bahya thus alludes to the composition attendant 
upon the tridimensionality of physical objects; and he does so although he could 
have made the more straightforward point that the elements are compounded of 
matter and the forms of each of the four elements. In his third piece of evidence 
for composition in the universe, Bahya refers, then, to both kinds of composi- 
tion — the composition of matter and form, and the composition of tridimension- 
ality — which had been adduced by Philoponus in the supporting arguments to 
his proof of creation from finite power 263 and which subsequently appeared in 
Kindi. 2 '" 1 

Bahya 's proof of creation from composition has, in fine, woven together the 
following strands: the argument that combination implies creation since the parts 
precede Ihc whole; the consideration that combination implies design; the con- 
sideration that the combination specifically of contrary qualities implies an exter- 
nal constraining agent; and illustrations of combination which are employed by 
Philoponus and reappear in Kindi. 



,2 Cf. Wolfson. Crescas, pp. 582-585. 
''Above, p. 92. 
Above, pp. 111-113. 



VI 



Arguments from the 
Concept of Particularization 



1. Inferring the existence of God from creation 

The standard Kalam procedure for proving the existence of God was to establish 
the creation of the world and then infer, from creation, the existence of a creator 
identified as the deity. The procedure has been termed the Platonic mode of 
proving the existence of God because of a passage in the Timaeus. Plato there 
offered a brief argument for the world's having been "generated and having begun 
at a first point," 1 whereupon he deduced the existence of a creator on the grounds 
that "what comes into existence must perforce come into existence through some 
cause." 2 Apparently Plato regarded the proposition that nothing can come into 
existence without "some cause" as self-evident. Galen's Compendium of the Timaeus, 
a text that happens to be known from a medieval Arabic translation, notes as 
much; Plato, writes Galen, put forward the proposition in question "without . . . 
demonstration, because it is something manifest to the intellect." 3 A number of 
Islamic and Jewish thinkers agree, either explicitly or implicitly, that the propo- 
sition is self-evident and that the inference of a creator from creation requires no 
justification. 

Ghazali, for example, advances an ostensible argument to show that what 
comes into existence has a cause of its coming into existence. 4 He adds, however: 
"In reality, the foregoing ... is not an argument . . . at all . . . but merely an 
explication of the terms generated and cause" ; for as soon as anyone "compre- 
hends the meaning of the terms, his intellect will necessarily affirm that whatever 
is generated has a cause." 5 Fakhr al-DIn al-RazI reports: "The need of what comes 

^Timaeus, 28B. The argument is not related to the medieval Arabic arguments. 
2 Timaeus. 28C. See above, Chapter I, n. 6. 

'Galen, Compendium Timaei Platonis, cd. P. Kraus and R. Walzcr (London, 1951). Arabic text, 
p. 4. Cf. also Corpus Hermelicum, cd. A. Nock and A. Festugiere (Paris, 1945-1954), XIV, 6. 
4 See below, p. 162. 

5 al-lqtisad fi ai-I' tiqad (Ankara, 1962), pp. 25-26. Cf. also Ghazali, al-Risala al-Qiidsiya, cd. 
A. Tibawi as Al-GhazaU's Tract on Dogmatic Theology (London, 1965), Arabic text, p. 16', English 
translation, p. 34. 

154 



Particularization Arguments 



155 



into existence for an agent" was considered by certain Kalam sheikhs, though 
not by everyone, to be an item of "necessary" and "immediate" knowledge. 6 
TusI has a similar report. He writes: In the view of the "later Kalam thinkers 
. . . the judgment that whatever conies into existence must inescapably have an 
agent bringing it into existence is immediate knowedge requiring no proof. . . ." 7 
The position taken by Ghazali and recorded by RazT and TusI is endorsed by ljl 
and his commentator. "The immediacy of the intellect." they assert, "testifies" 
that "whatever comes into existence has an agent bringing it into existence." 8 

Several figures who stand on the periphery of the Kalam, while not explicitly 
characterizing the proposition as self-evident, do treat it as such. Kindi and Joseph 
ibn Saddiq ask whether the cause bringing a thing into existence might not be 
the thing itself; but as for the principle that anything coming into existence docs 
have some cause or other, it is treated by them as beyond question. 9 Bahya too 
takes for granted that everything coming into existence has a cause. After pre- 
senting his proof for creation he does feel called upon to state that it is "impossible 
for anything to produce itself."'" But once the statement has been made, he can 
conclude: "We know by the testimony of healthy intellects" that if the world has 
come into existence, "something outside the world created and brought the world 
into existence."" In other words, as soon as the impossibility of something's 
producing itself is recognized, the human intellect testifies that what comes into 
existence is brought into existence through a cause from without. The more 
comprehensive principle that things coming into existence do have some cause 
or other is, Bahya implies, self-evident. 

For each of the authors referred to thus far, the proposition that what comes 
into existence has a cause is, then, self-evident; and each of them accordingly 
can — and docs — state his proof or proofs of the creation of the world and infer 
the existence of a creator forthwith. Kalam writers who, on the contrary, did not 
regard the proposition as self-evident had to justify their inference of a creator 
from creation. Two lines of reasoning were pursued. Either an analogy was drawn 
between the coming into existence of objects and events within the world and 
the coming into existence of the world as a whole; or else recourse was had to 
an argument turning on the concept of particularization or the kindred concept 
of tipping the scales. 

Saadia, in the passage that is germane, does not spell out his position but he 
seems to pursue the former line. He has an imaginary interlocutor trace the 
judgment that "there is nothing made without a maker" to what is "testified" in 

6 A'. ul-Arha'm (Hyderabad. 1934), p. 89. 
'Gloss to RazI, Muliassal (Cairo, 1905), p. 106. 
"Mawaqif (Cairo, 1907), VIII, p. 3. 

"Kindi, Rasa it (Cairo, 1950), I, pp. 123, 207; Joseph Ibn Saddiq, ha-'Olam ha-Qatan, ed. S. 
Horovitz (Breslau, 1903), p. 49. 

"\il-Hidaya (Ifobol ha-Lebabol), cd. A. Yahuda (Leiden, 1912), I, 6. 
"Ibid., I, 7(3). 



156 



Particularization Arguments 



the realm of "the perceptible." The interlocutor's meaning, and probably Saadia's 
as well, is that in the realm of sense experience things coming into existence are 
dependent on a cause; and by analogy or induction, things coming into existence 
outside the realm of human experience, including the world in its entirety, can 
also be presumed to be dependent on a cause. 12 Maturidi (d. 944) speaks more 
directly and clearly. "Building, writing, and ships," he writes, "testify to what 
we have said. For they can come into existence only through an existent agent, 
and the present instance," the world's coming into existence in its entirety, "must 
be similar" ; human artifacts are known empirically to come into existence through 
an agent, and by analogy, the world as a whole — given its creation — must also 
be the work of an agent. 13 The reference by Saadia and Maturdil to what is 
testified to empirically contrasts nicely with the previous references to the testi- 
mony of the intellect. M Maturidi gives another version of the line of thought we 
are considering when he argues: "In the world as perceived (shdhicl) 15 nothing 
exists which combines or separates by itself. It follows that such must have been 
true of the state of the world which cannot be perceived (gha'ib)." 16 That is to 
say, just as nothing now coming into existence through the combination and 
separation of atoms can dispense with a cause, so too at the moment when the 
world as a whole came into existence, the elements constituting the world can- 
not — whether they themselves came into existence or already existed — have com- 
bined and separated spontaneously and without a cause. In BaqillanI the argument 
runs: "Writing unquestionably requires a scribe; drawing, an artist; and building, 
a builder. Should anyone tell us that something written has come about without 
a scribe, or something molded without a molder, or weaving without a weaver, 
we would not doubt the speaker's ignorance." But the world is of a "more subtle 
and wondrous artisanship." Inasmuch as the world is known to "have come into 
existence," it must, by analogy and with even greater certainty, have a "maker." 17 
A similar formulation is found in Baghdad! 18 and, very briefly, in Bazdawl. 19 

'Abd al-Jabbar adds a degree of analysis that raises the discussion from the 
level of analogy to what may be called the level of induction. "In the realm of 
what is perceptible (shdhid)" he reasons, "our operations . . . stand in need of, 
and are dependent on, us." So much is obvious from the circumstance that human 

,2 K. al-Amanat wa-l-l'tiqddat, ed. S. Landauer (Leiden, 1880), I, 2, p. 39; English translation, 
with pagination of the Arabic indicated: Book of Beliefs and Opinions, trans. S. Rosenblatt (New 
Haven, 1948). 

"K. al-Tawhid (Beirut, 1970), p. 18. 

'"Above, pp. 154-155. 

l5 The root of the word is the same as that in the term to testify. 
I6 K. al-Tawhid, p. 17. I have corrected bi-ghayrihi to bi-ghu'ib. 

n K. al-Tamhid (Beirut, 1957), p. 23. BaqillanT may well have in mind the argument from analogy 
in Ashari, Luma', I, 4. A Ideological motif, such as will be discussed in the next chapter, also seems 
to be present. 

iS K. UsCd al-Din (Istanbul, 1928), p. 69. 

"K. Usui al-Din. cd. H. Linss (Cairo, 1963), p. IX. 



Particularization Arguments 



157 



operations "occur in conformity with our intention." Furthermore, the aspect of 
human operations rendering them dependent on the human agent can be nothing 
other than the fact of their "having come into existence." That is obvious from 
the circumstance that "what occurs in conformity with our intent" is precisely 
"their coming into existence"; since what occurs in conformity with the agent's 
intent is precisely the coming into existence of an action or operation, coming 
into existence and nothing else must be the aspect of actions and operations 
rendering them dependent on the human agent. Now by the Kalam rule of induc- 
tion, whenever a primary characteristic or "ground" ('ilia) is, as far as experience 
goes, invariably accompanied by another characteristic, the connection between 
the characteristics can be presumed to obtain where experience does not pene- 
trate. 20 In the instance at hand, the "ground" is the characteristic of having come 
into existence, and the accompanying characteristic is dependence on an agent. 
'Abd al-Jabbar therefore affirms that "whatever has in common with our oper- 
ations [the characteristic oil having come into existence, must likewise have in 
common with them [the characteristic of] requiring an agent." Hence once the 
world is known to have come into existence, an agent must be posited who brought 
it into existence. 21 The argument that anything coming into existence requires a 
cause is formulated in almost the same terms by a Jewish Kalam work attributed 
to Abraham ibn Ezra, 22 and by two Karaite authors, Joseph al-BasIr and Jeshua 
b. Judah, who are known to have stood under the influence of 'Abd al-Jabbar's 
school. The Karaite authors explain: The "ground" ( 'ilia) in human acts rendering 
them dependent on man for their existence is the fact of their "having come into 
existence." But when a given ground is invariably observed to be accompanied 
by another characteristic, whatever possesses the former must — even in areas 
beyond the scope of observation— possess the latter. In the instance at hand, the 
world has in common with human actions the characteristic of "having come into 
existence." It must consequently have in common with human actions the need 
for "an agent bringing it into existence." 2 ' 1 

The foregoing quotations from Saadia, Maturidi, BaqillanI, 'Abd al-Jabbar, 
and 'Abd al-Jabbar's Jewish followers, all look like variant formulations of a 
single theme, though some are expressed more technically than others, and I have 
termed them a single line of reasoning and contrasted them with the earlier 
position according to which the dependence of things coming into existence on 
a cause is self-evident. Fakhr al-DIn al-Razi viewed the matter differently. He 
drew a sharp distinction between "two approaches" to the inference of a creator 

2 "Cf. J. van Ess, Die Erkenntnislehre des 'Adudaddin ul-lct (Wiesbaden, 1966), pp. 361, 381- 
391. The problem of determining which is consequent and which is ground is obvious. 
21 Sharh al-Ustil (Cairo, 1965), pp. 118-119. 
"Published in Kerem Hemed. IV (1839), 3. 

■"Joseph al-Basir, in P. Frankl, Bin Mu'tazililischer Kalam aus dent 10. Jahrhundert (Vienna, 
1872), p. 21; and in M. Schreincr, Studien ueber Jeschu'a ben Jehuda (Berlin, 1900), p. 39, n. I . 
Jeshua, in Schreincr, pp. 38-39. 



158 



Particularization Arguments 



from creation; and some of the argumentation that has been quoted is subsumed 
by him under the first approach, while other argumentation is designated as the 
second. 

RazI writes: Certain Kalam thinkers who inferred a creator from creation main- 
tained that "the need of what comes into existence for an agent is [a piece of] 
necessary knowledge." To "prove" their thesis they pointed out that "anyone who 
sees a building erected or a castle upraised would know necessarily that the 
structure had a builder and maker; and, indeed, should someone allege that the 
building might come into existence without a maker and builder, he would be 
judged mad. We thus recognize the proposition to be [a piece of] immediate 
[knowledge]." 24 RazI is not saying that for the writers in question the coming 
into existence of the world is analogous to the coming into existence of a building, 
and just as we perceive the building to be the work of a builder, so by analogy 
should we judge creation to be the work of a creator. The step from the coming 
into existence of an effect to the existence of a cause and, more specifically, from 
creation to a creator is represented, in Razl's account, as "immediate" knowledge; 
and the example of a building and a builder serves not to prove the proposition's 
truth but rather to "prove" its immediacy and sclf-cvidcncc. The example serves 
as a touchstone or mental experiment for eliciting what the intellect apprehends 
as self-evident truth. Men might not — although it sounds a bit paradoxical — 
realize at once that creation immediately entails a creator, since the creation of 
the world is an unfamiliar event. By testing himself with more familiar events 
like the coming into existence of human artifacts, a person may — such is the 
burden of Razl's account — more easily recognize that the step from something's 
coming into existence to an agent bringing the thing into existence is self-evident. 

That, according to RazI, is one approach to the inference of a creator from 
creation. The proposition that what comes into existence requires a cause is taken 
as self-evident and the example of human artifacts serves to elicit its self-evi- 
dence. The second approach, he continues, is pursued by "most of the Mu'tazilite 
sheikhs." It treats the proposition that things coming into existence are dependent 
on a cause not as self-evident but as "something to be proved"; and the proof 
resorted to turns out to be the argument by induction met in 'Abd al-Jabbar. 25 
The reasoning is given by RazI as follows: We perceive from our own experience 
that "the human agent brings his own actions into existence." We discern, more- 
over, that human "actions stand in need of us precisely because they come into 
existence after not having existed." The "ground" for their standing in need of 
an agent is thereby revealed to be nothing other than the fact of their "coming 
into existence." When a given ground is observed to be invariably accompanied 
by another characteristic, whatever possesses the former will possess the latter. 
The "world" as a whole does possess the critical ground of having "come into 

24 AT. al-Arba'in, p. 89. 

"RazT refers to other, less familiar Mu'lazilites. 



Particularization Arguments 



159 



existence." The world must consequently possess the accompanying characteristic 
of "standing in need of an agent." 26 

RazI expects his readers to keep in mind the differing stands taken by Mu'tazilitcs 
and Ash'aritcs on the nature of human actions. Mu'tazilitcs held the common- 
sense view that men perform or "create their own actions," while the Ash'aritcs 
totally negated natural causation and subscribed to an occasionalistic scheme 
wherein God, not man, creates every human action. BaqillanI and BaghdadI were 
members of the latter school, and Maturidl belonged to a cognate school. There- 
fore when they state that things coming into existence require a cause and creation 
requires a creator, just as a building requires a builder, they cannot, RazI surmises, 
be genuinely thinking of the erection of a building by a human builder; for they 
believed that buildings and other artifacts are never fashioned by human hands. 27 
When these writers compare the dependence of a building on a builder with the 
dependence of creation on a creator, they should accordingly be read not as 
drawing an analogy between a human artisan and the divine artisan, but merely 
as indicating how the self-evidence of the need of what comes into existence for 
a cause can be elicited. A true analogy or induction from human action to divine 
action could only be offered by a Mu'tazilite. And 'Abd al-Jabbar, a Mu'tazilite, 
docs unmistakably offer an argument of the sort. He analyzes human acts, under- 
takes to isolate the characteristic rendering them dependent on the human agent, 
and generalizes from human actions to everything containing the critical char- 
acteristic. RazI, an Ash'aritc, naturally enough rejects the Mu'tazilite argument 
by induction. He insists that if the existence of God is to be inferred from creation, 
the principle that everything coming into existence requires an agent must be 
accepted as a self-evident truth. 21 * 

RazI, then, demarcates the boundary between those who treat the inference of 
a creator from creation as self-evident and those who defend the inference by 
analogy and induction in a different place from where the obvious sense of the 
texts would locate it. But whatever the merits of his interpretation, one line of 
reasoning in support of the inference of a creator from creation, should the 
inference not be regarded as self-evident, did take the form of an analogy or 
induction. The second line of reasoning in support of the inference brings the 
concept of particularization into play. Since the two lines are nowise incompat- 
ible, several of the writers to be cited — Maturidl, BaqillanI, and BaghdadI — 
could deploy them side by side. 

The particularization mode of argument searches for instances in the universe 
where, it understands, a given alternative has been selected over other, equally 
possible alternatives; and it submits that the arbitrary selection it discovers implies 

2, 'K. al-Arba'in, pp. 89-90. Followed by Iji, Mawaqif, VIII, p. 3. 

27 The building, which we arc certain has a builder, is in other words fashioned by God. 

2K K. al-Arba'in. p. 90. Juwayni, K. al-Shamil (Alexandria, 1969). p. 280. struggles with the 
question how an analogy might be drawn to human actions if men arc not truly the authors of their 
actions. 



160 



Part kularizat ion Arguments 



a particularizing agent or a particularizing factor. In Maturldi, the term parti- 
cularization is absent, but the thought is unmistakable. Maturldi builds on his 
proof of creation and writes: If the world came into existence spontaneously, "no 
one time would be more appropriate for its coming into existence than another." 
The world's having come into existence at a given specific time rather than another 
therefore "proves" that the world came into existence "through something else," 
through an agent who arbitrarily selected the time for it to appear. 29 Maturldi 
might have said, as subsequent writers do, that a particularizing agent must have 
selected out a particular time for the world to come into existence. 

BaqillanI too has an argument from creation to a particularizing agent; he, 
however, looks not at the coming into existence of the world as a whole but at 
the coming into existence of the world's myriad parts, and he is able to draw his 
inference only by presupposing the Ash'arite denial of natural causation. In his 
proof for creation, BaqillanI had established that accidents and bodies — the latter 
being composed of inert atoms to which accidents are conjoined — must have all 
come into existence. 30 After completing that proof, BaqillanI proceeds: Things 
coming into existence are "similar to one another"; an accident carrying any 
quality is, in other words, exactly like other accidents carrying the same quality, 
and a physical object consisting of a number of atoms to which a set of accidents 
is conjoined is exactly like other physical objects consisting of the same number 
of atoms and the same set of accidents. Yet "despite . . . being similar to one 
another," things that come into existence make their appearance at different times, 
"some earlier . . . than others," and "some later." The cause of a thing's coming 
into existence at an earlier or a later time cannot be the thing "itself and its 
genus"; "for should a thing come into existence earlier or later by virtue of itself, 
everything of the same genus would come into existence at the [same] earlier |or 
later] time." BaqillanI ignores the commonsense explanation for some objects' 
coming into existence earlier and some, later — namely, that natural forces within 
the universe, which operate in accordance with natural laws, determine the time 
when things come into existence. He can as a consequence conclude: The fact 
that a given physical object does come into existence earlier or later than some- 
thing exactly like it "is proof that it has an agent rendering it early [or late] and 
assigning it a definite span of existence in conformity with the agent's will." The 
aggregate of objects making up the physical world must thus depend on an 
agent — and BaqillanI will subsequently explain why there can be only one — 
who arbitrarily assigns each object a time for it to exist. 31 Although BaqillanI 
does not expressly speak here of a particularizing agent who brings the world 
into existence, he employs the terms "particular" and "particularizing" a few 
lines afterwards in a parallel context. 32 The argument that each physical object 

2 *tf. al-Tawhid, p. 17. 
30 Cf. above, p. 136. 
3I A\ al-Tamhld, p. 23. 
"Ibid., and below, p. 178. 



Particularization A rguments 



161 



coming into existence must have the time for its emergence assigned by an agent 
is advanced by BaghdadI as well; and BaghdadI docs expressly call the latter a 
"particularizing agent" (mukhassis). 33 

Juwaynl also has a particularization argument, but he, like Maturldi, applies 
it to the world as a whole and thereby frees it again of the occasionalistic burden 
with which BaqillanI and Baghdad! loaded it; if the world has come into existence 
as a whole there are — even on a naturalistic picture of the universe — no natural 
forces operating in accordance with natural laws which might have determined 
the moment for it to come into existence. Juwaynl's reasoning goes: The creation 
of the world has been demonstrated, and whatever comes into existence has an 
equal possibility of "existing . . . and not existing." The world, moreover, might 
have come into existence at different "possible" times. From the world's having 
come into existence at a given time rather than "continuing in a state of nonex- 
istence," the "intellect immediately judges that the world requires a particular- 
izing agent who selected out existence for it" at the time when it came into 
existence. 34 Juwaynl's use of the phrase "the intellect immediately judges" is 
revealing. The particularization argument was called into play by Kalam thinkers 
in order to furnish underpinning for the inference of a creator from creation and 
not leave it an item of self-evident knowledge. But, Juwaynl recognizes, the step 
from the selection between equal possibilities to a particularizing agent who 
makes the selection must itself ultimately be taken as self-evident. 

The thought that creation points to a particularizing agent who selected out the 
particular moment for the world to come into existence is employed by Bazdawl, 35 
a follower of Maturldi, and by Ghazali; 36 and it is recorded by Judah Hallcvi in 
his summary of Kalam doctrines. 37 

The passage quoted from Juwaynl merits further comment, since it reveals, in 
fact, not one, but two intertwined motifs. Juwaynl is saying both that a particu- 
larizing agent must have selected out a moment for the world to emerge in 
preference to other times when the world might have emerged; and that a parti- 
cularizing agent must have selected out existence for the world in preference to 
nonexistence. 38 The former is the motif met already in Maturldi, but the latter is 
a new motif, which must have been suggested to Juwaynl by Avicenna's analysis 
of the concepts possibly existent and necessarily existent. 39 Juwaynl realized that 
the two motifs are distinct. 40 And in a composition that was designed as a mere 
outline of his thought and not as a full-fledged theological work he restricts 
himself to the second of the two. Given creation, he explains there, the world 



"K. Usui al-Din. p. 69. 

M K. al-lrshad (Cairo, 1950), p. 28; cf. K. al-Shamil, pp. 262-265, 267. 
"K. Usui al-Dln, p. 18. 

"'al-Risdla al-Qudsiya, Arabic text, p. 16; English translation, p. 34. 
"Kuzari, V, 18(3). 

"K. al-Shamil, pp. 263-265. shows how Juwaynl came to combine the motifs. 
19 Cf. below, p. 290. ""Below, p. 177. 



162 



Particularization Arguments 



must have a creator. For "things coming into existence" arc "possibly existent 
(ja'iz al-wujud), inasmuch as either their existence or — in preference to their 
existence — their continued nonexistence can be supposed. When they are parti- 
cularized in [or: by] actual possible existence (wujiid mumkin) they stand in need 
of a particularizing agent." 41 

Echoes of Avicenna are more audible in a variation of the particularization 
argument offered by Ghazali, Fakhr al-DIn al-RazI, and Ijl. Avicenna had insisted 
that when something possibly existent becomes actual, actual existence must be 
"differentiated out" for the object over nonexistence. 42 It was not Avicenna's 
intent that the world is anything but eternal, and he certainly did not have in 
mind a temporal differentiation out of actual existence for the world as a whole. 
Ghazali and the others, however, turn Avicenna's analysis to their own purpose. 
They arc once more justifying the inference of a creator from creation. They 
focus exclusively on the selection of existence for the created world in preference 
to nonexistence; and they now employ not the language of particularization, but 
kindred language, that of tipping the scales. Everything coming into existence, 
they write, plainly is "possibly existent" prior to actually existing; for if a thing 
were "impossible," it would never exist, whereas if it were "necessarily existent 
by virtue of itself," it would always exist. But anything whose existence is possible 
is, in itself, equally capable of existing and not existing. Something of the sort 
can, therefore, enter the domain of actual existence only through an agent that 
"tips the scales (murajjih) in favor of its existing." The world is known, thanks 
to Kalam proofs of creation, to be an object that came into existence, and hence 
an object that was possibly existent before actually existing. When the world 
entered the domain of actual existence an agent must, then, have tipped the scales 
in favor of its existence. 43 

Such were the Kalam arguments supporting the inference of a creator from 
creation. A supplementary detail was sometimes provided both by those who 
treated the inference of a creator from creation as self-evident and those who 
offered arguments to support the inference. 

The rhetorical question is posed: Granted that the world must have had a cause 
bringing it into existence, why might not the world itself have been the cause? 
Why cannot an object bring itself into existence and select out the particular 
moment for its own emergence? The most common response goes back at least 
to Proclus' commentary on Plato's Timaeus. When expatiating on Plato's state- 
ment that "what comes into existence must perforce come into existence through 



"Textes apologetiques de Guwaini (Luma'), ed. M. Allard (Beirut, 1968), pp. 128-129. 

"Najat (Cairo, 1938), p. 226; Shifa : llahiyal, ed. G. Anawati and S. Zayed (Cairo, 1960), p. 
39; French translation, with pagination of the Arabic indicated: La Metaphysique du Shifa', trans. G. 
Anawati (Paris, 1978). 

"Ghazali, al-lqlisad, pp. 25-26; Raz.I, K. al-Arba'in, p. 86, together with p. 71; cf. idem, 
Muhassal, p. 106, together with pp. 53-54; Ijl, Mawaqif, VIII, p. 3. 



Particularization Arguments 



163 



some cause," Proclus wrote: "Nothing can possibly bring itself into existence; 
for if it did, it would exist before corning into existence" which is absurd. 45 
Saadia, Maturidi, Baghdad!, Bahya, and Joseph ibn Saddiq argue in the same 
vein. Before something has come into existence, they point out, it surely cannot 
be described as the agent bringing it into existence; for it docs not yet exist, and 
what docs not exist cannot function as an agent. But after a thing has already 
come into existence it again cannot be described as the agent bringing it into 
existence, since it already exists and no longer needs to be brought into existence. 
There is thus no time at which a thing can be described as bringing itself into 
existence; and the world cannot have been brought into existence by itself. 46 'Abd 
al-Jabbar says more or less the same in different words. Nothing, he contends, 
can bring itself into existence, since only what has power can bring things into 
existence, and "what has power over anything must precede its effect. If the agent 
bringing a body into existence were the body itself, it would have to have power 
while still nonexistent," which is impossible. 47 

Kindi and Ibn Hazm introduce the following consideration: To state that an 
agent brings something into existence is to affirm, in effect, that two entities 
exist, one of which acts upon the other. The supposition that something brings 
itself into existence would, hence, be tantamount to supposing that a single thing 
is two distinct things. Since the supposition is self-contradictory and absurd, 
nothing, including the world, can bring itself into existence. 48 

Additional considerations were advanced which, we should probably agree, 
fall somewhat below the threshold of philosophy. Baqillani reasons: The world 
contains death; but what contains death cannot be a creator; therefore the world 
cannot be a creator. 49 Saadia and MaturidI reason: If the world had the capability 
of creating itself, it would likewise have had the capability of desisting from 
creating itself. To speak, however, of the world's desisting is to imply that the 
world existed, whereas to speak of its desisting from creating itself would be to 
imply that the world had never been created and never existed. The notion of the 
world's desisting from creating itself is thus sclf-conlradictory and impossible. 
By definition, the possible is such that when it is assumed to exist, nothing 



Above, n. 2. 

"Commentary on Timaeus, ed. E. Dichl (Leipzig, 1903-1906), 1, p. 260. lines 8-9; French 
translation, with pagination of the Greek indicated: Commentaire sur le Timee. trans. A. Festugicre 
(Paris. 1966-1967). 

46 Saadia, K. al-Amanal. I, 2, p. 38; MaturidI, K. al-Tawhid, p. 18; Baghdad?, A". Usui al-Din. 
p. 69; Bahya, al-HUlaya (Hobot ha-Lcbabot), I, 5; Ibn Saddiq, ha- 'Olam ha-Qatan. p. 49. Cf. below, 
n. 48. The argument was known to Hume, Treatise of Human Nature, I, iii, 3. 

47 ,S7m/7i al-Usul, p. 1 19. 

4 *Kindi. Rasail, 1, pp. 123-124; Ibn Hazm. K. al-FasI ft al-Milal (Cairo, 1964). I, p. 18. At 
first Kindi seems to be arguing as Saadia, Baghdad!, and the others did, but then he slides into this 
argument. 

4 "A'. al-Tamhid, p. 24. Cf. Maturidi. K. al-Tawhid, p. 17. 



164 



Particularization Arguments 



impossible results therefrom. 50 Since the hypothesis that the world had the capa- 
bility of creating itself leads to the impossible notion that the world could have 
desisted from creating itself, since the hypothesis docs result in an impossibility, 
it fails to satisfy the definition of the possible. That the world might have created 
itself is therefore not possible." 

Saadia, to give a final instance, writes: When a thing exists, it is stronger than 
when it does not exist. The world in its state of existence is too weak to create a 
world. A fortiori, the world would not have been strong enough to create a world 
before existing. 52 

To recapitulate: The standard Kalam procedure for proving the cxislcr.ee of 
God was to infer a creator from creation on the grounds that what comes into 
existence must have a cause. Some Kalam authors treated the proposition that 
things coming into existence must have a cause bringing them into existence as 
self-evident. Of those who did not treat the proposition as self-evident, some 
supported the inference of a creator from creation through an argument by analogy 
or induction. And some supported it through an argument turning on the concept 
of particularization or the kindred concept of tipping the scales. Once satisfied 
that the coming into existence of the world does require a cause, a number of 
Kalam writers appended a supplementary detail. They explained why nothing 
can bring itself into existence and why the creator of the world cannot have been 
the world itself. 

Adherents of the Kalam do not stop here. They proceed to argue that the cause 
of the existence of the world must be eternal; several undertake to prove that 
what brought the world into existence cannot be a nature or necessary cause; 
invariably they argue that the cause of the world must be one and incorporeal and 
that it must possess power, knowledge, life, and will. Unfortunately, Kalam 
thinkers do not make clear which of the foregoing they consider to be part of the 
concept of God, indispensable conditions such that nothing can merit the des- 
ignation deity without them; and which are ancillary attributes, not integral to 
the very concept of deity. If an attribute is deemed integral to the concept of the 
deity, a proof of .the existence of God will not be complete until the cause of the 
universe is shown to possess it. If, for instance, 'God' is taken to mean the 'single 
incorporeal cause of the world,' the existence of God will not have been estab- 
lished until the world is shown to have a cause that is one and incorporeal, whereas 
if 'God' means merely the 'cause of the world,' a proof of the existence of God 
can be achieved without arguments for unity and incorporeality. 

Of the attributes that have been mentioned, unity, especially, might be expected 
to be integral to the Kalam concept of the deity; the unity of God was so central 
a Kalam doctrine that 'establishing the unity [of God]' (tawhid) was the Kalam 



'"Aristotle, Prior Analytics I, 13, 32a, 18-20. 

5, Saadia, K. al-Amanal, p. 38; Maturidi, K. al-Tawhid, p. 18. 

32 A. al-Amanal, p. 37. 



Particularization Arguments 



165 



term for 'natural theology.' The texts dealing with the unity of God are, as 
happens, ambiguous. Writers propose to demonstrate that "God is one," 5 '' and 
by expressing themselves in that way may appear to grant that God can be known 
to exist before he is known to be one. And yet in the most popular proof for 
unity, the pivotal thought is that the hypothesis of two deities cannot be squared 
with what is meant by 'God'; and there the presupposition would appear to be 
that unity is after all an indispensable condition for the deity. The writers who 
follow their proof for the existence of a creator with an argument establishing 
that the creator cannot be a nature or necessary cause undoubtedly consider this 
characteristic at least, the characteristic of not being a nature or necessary cause, 
to be an indispensable specification for the deity. And a few other attributes also 
are represented as indispensable specifications. 54 But as for the remainder, includ- 
ing unity, it is hard to determine which the generality of Kalam writers did, and 
which they did not, understand to be specifications for the deity, and where, 
consequently, their proofs of the existence of God end and their arguments for 
the ancillary attributes begin. 

At any rate, virtually all who infer a creator from creation add that not merely 
docs the existence of the world depend upon a cause; it depends upon a first 
eternal cause. The stock explanation is that an infinite regress would otherwise 
ensue. As Baghdad! puts it: "If the creator himself came into existence, he would 
stand in need of an agent to bring him into existence. If the latter too came into 
existence, he in turn would stand in need of a third agent. And the scries would 
regress (yatasalsal) infinitely, which is absurd." To avoid an infinite regress, the 
"creator" of the world — or, to be more precise, the ultimate creator — must be 
judged "eternal." 55 The argument for the eternity of the creator from the impos- 
sibility of an infinite regress is given by Baqillani, 56 'Abd al-Jabbar, 57 Juwaynl, 58 
Ghazali, 59 Bahya, 60 Joseph al-Basir, 61 Judah Ilallcvi, 62 Abraham ibn Ezra," 
Fakhr al-DIn al-RazI, 04 and iji. 65 It contrasts slightly with an argument employed 
by Aristotle and his medieval disciples. Aristotle and the medieval Aristotelians 
reached a first cause, not a first creator, of the world with the aid of a parallel 



"'Abd al-Jabbar, Shark al-Usul, p. 277; Juwaynl, K. al-lrshad, p. 53; Ghazali, al-Iqlisad, p. 
73; iji, Mawuqif, VIII. p. 39. 
vl See below, p. 167. 
"A". Usiil al-Din. p. 72. 
56 A'. ai-Tamhid. p. 25. 
"Sharhal-Usfd, p. 181. 

5 *A. al-lrshad. p. 32; K. al-Shamil. pp. 617-618. 
^al-Iqlisad, p. 35. 

m al-Hiddya (Hobot ha-Lcbabol), 1, 6; 10. 
6l Frankl, Ein Mu'tazilitischer Kalam, p. 24. 
62 Kuzari, V, 18(4). 
"Kerem Hemed, IV, p. 4. 

64 A'. al-Arba'in, p. 92 (in the name of the "sheikhs" of the Kalam). 
"Mawaqif, VIII, pp. 4-15. 



166 



Particularization Arguments 



consideration, the consideration that an infinite regress of causes is impossible. 
But their grounds for ruling out an infinite regress of causes were that the nature 
of causation precludes causes' running to infinity; 66 whereas the adherents of the 
Kalam rule out an infinite series of agents bringing the world into existence at a 
particular moment on the grounds that an infinite regress of any sort is impossible. 
The difference may sound insignificant. Yet Ghazali made a good deal of it and 
undertook to show that the Aristotelian approach to the infinite regress embodies 
an inconsistency and as a consequence is, unlike the Kalam approach, untenable. 67 

Kalam argumentation has arrived at a first eternal cause of the creation of the 
world. BaqillanI, 'Abd al-Jabbar, Juwaynl, and Jeshua b. Judah, furnish an argu- 
ment establishing that the first cause cannot be a "nature" or a "[necessary] cause" 
('ilia), but must be a "voluntary agent." The reasoning goes: The cause of the 
world is known to be eternal. Both a nature and a necessary cause operate in an 
unvarying manner; and if an eternal nature or eternal necessary cause had acted 
to produce the world, it would have acted as long as it existed, and the world 
would likewise be eternal. But the Kalam proofs of creation have demonstrated 
that the world was created. The world therefore cannot have been produced by a 
nature or a necessary cause, and must instead be the work of a "voluntary agent." 68 

A variety of grounds, all highly dialectical, were adduced to establish the unity 
of God. Most popular was (a) the argument from mutual interference (tamanu'), 
an argument whose provenance apparently was Greek. The author of the I lermctic 
Corpus contended that if more than one deity existed, "rivalry" would beset 
them. 59 And John of Damascus defended the unity of God with the argument 
that if "several gods" existed, "conflict" between them could not be avoided, 
and the world would be "broken up and utterly destroyed." 7 " For their part, the 
Islamic writers traced the argument from mutual interference not to Greek sources 
but to Scripture. 

Quran 21:22 reads: "If gods other than God were in them [i.e., in the heavens 
and earth], both [the heavens and earth] would fall into ruin." And Quran 23:91 
reads: "There is no god along with him; else each god would assuredly have 
championed what he created, and one would have overcome the other." Ash'ari 
spells out the intent of the Quranic verses: On the hypothesis of two creators, 
one of the two might will something while the other willed the contrary. For 
example, -one might "will to have a man live, while the other willed to have him 
die." Should such occur, it would not be possible for the will of both to be 
accomplished; nor for the will of neither to be accomplished; nor again for the 



66 See below, p. 337. 
67 Sce Appendix A. 

68 Baqillani, K. al-Tamhld. pp. 34-35; 'Abd al-Jabbar, Sharh al-Usul, p. 120; Juwaynl, K. al- 
hshad, pp. 28-29; Schreincr, Studien ueber Jeschu'a ben Jehuda, p. 38. Baghdad!, K. Usui al- 
Din, p. 69, has a different argument showing that the creator is not a "nature." 

''''Corpus Hermelicum, XI, 9. 

m De Fide Orthodoxa. I, 5. 



Particularization Arguments 



167 



will of one to be accomplished, while the will of the other was frustrated. Clearly, 
"what both will could not conceivably be accomplished; for a body cannot con- 
ceivably be alive and dead at the same time." Nor might "what neither wills be 
accomplished." The reason therefor is that if the will of neither creator were 
accomplished, both assumed creators would "perforce be powerless, whereas," 
Ash'ari postulates, "what is powerless cannot be God or eternal." Later writers 
would here interpose an additional, logical reason why it would be impossible 
that the will of neither supposed deity should be accomplished; if the will of 
neither were accomplished, the body would be neither dead nor alive; and the 
law of the excluded middle would be violated. 71 Finally, writes Ash'ari, it is 
impossible that the will of one of the two assumed divine creators should be 
accomplished while the will of the other is frustrated. For if "what one of the 
two assumed creators wills is accomplished to the exclusion of the other." the 
former alone would be God, seeing that the other would be powerless, and 
"anything powerless cannot be God or eternal." Thus, no more than one divine 
creator can, Ash'ari concludes, exist. 72 The argument from mutual interference 
is employed by Saadia 7 "' and Maturidi, 74 who like Ash'ari were active in the 
early tenth century; and it undoubtedly was already a stock argument by the end 
of the previous century. 75 

When BaqillanI subsequently gave his version, he put Ash'ari's remark con- 
cerning power and eternity a trifle more explicitly. "Powcrlcssncss," BaqillanI 
comments in the course of restating Ash'ari's argument, "is a mark of having 
come into existence," and for that reason "the eternal cannot be powerless." 76 
The presupposition in Ash'ari and, more explicitly, in BaqillanI — a presuppo- 
sition other Kalam thinkers employing the argument concur in — is that eternity 
and power are part of the irreducible concept of the deity, and nothing can be 
designated a deity without being eternal and powerful. Inasmuch as the unity of 
God is being demonstrated through the presupposition that eternity and power 
are integral to the concept of God, unity too is perhaps being construed as integral 
to the concept of God. 

The argument from mutual interference invites an objection, and the objection 
is taken up by 'Abd al-Jabbar. He presents the argument from mutual interference 
much as Ash'ari did, although the imagined conflict between the deities which 
he outlines is slightly different; he assumes not that one of the supposed deities 
wills life for a given body while the other wills death, but rather that one wills 

"Baghdad!. Juwayni, Shahrastani, and Ijl, in passages cited below, nn. 80. 81. 83, 84. 
72 K. al-Luma'. in Theology of al-Ash'ari, ed. and trans. R. McCarthy. (Beirut. 1953). 58. 
"K. al-Amanai, II, 3. p. 82. 
74 A'. al-Tawhid, p. 20. 

"The argument seems to be present in al-Muqammis (ninth century). Sec G. Vajda. "Le Problcmc 
dc I' Unite de Dieu d'aprcs al-Muqammis," Jewish Medieval and Renaissance Studies, cd. A. Altmann 
(Cambridge, Mass. 1967), p. 57. 

7< T. al-Tamhld. p. 25. 



168 



Particularization Arguments 



motion while the other wills rest. Should conflict of the sort occur, it would, 
'Abd al-Jabbar shows, be impossible for the will of both supposed deities to be 
accomplished; it would be impossible for the will of neither to be accomplished; 
and if the will of only one were accomplished, the one whose will was accom- 
plished would alone be the deity. 77 'Abd al-Jabbar concludes that only one deity 
can exist, whereupon he turns to the objection. 

The objection runs: Granted that two deities could not exist who will contrary 
effects, might not two deities "be wise and never interfere with each other?" 
Might there not exist two deities who always will identical events and always see 
their will accomplished? 'Abd al-Jabbar's response is that the argument from 
mutual interference requires not the "reality" of a conflict of wills, but merely 
the "supposition" (taqdir) or "possibility" (sihha) of a conflict. Since on the 
hypothesis of two deities the two would have the possibility, at least, of willing 
contrary effects, the question remains which of them would, in the event of a 
conflict, be able to execute his will. And the one possessing the ability to execute 
his will could alone be called the deity. 78 'Abd al-Jabbar may well have in mind 
the definition of the possible as what is such that "when it is assumed to exist, 
nothing impossible results therefrom." 79 His thinking would be that the hypoth- 
esis of two deities entails the possibility, at least, of their willing anything in 
their power, and hence their willing contrary events. The existence of two deities 
who will contrary events has, however, been shown to be an impossibility. The 
hypothesis of two deities is thus found to be an assumption from which something 
impossible does result, and the hypothesis fails to satisfy the definition of the 
possible. No more than one deity can, consequently, exist. 

The argument from mutual interference, together with the response to the 
objection that two deities might perhaps always will identical effects, is put 
forward by BaghdadI, 80 Juwaynl, 81 Joseph al-BasIr, 82 ShahrastanI, 83 and ijl. 84 
The argument without the objection and response to it is given by TaftazanI, 85 
and is alluded to by Bahya 86 and the Karaite Judah Hadassi. 87 



"Shark al-Usul, p. 278. 

ls Sharh al-Usiil, p. 283, with further elaborations. Cf. K. al-Majmu' ft al-Muhit bi-l-Taklif, ed. 
J. Houben (Beirut, 1965), pp. 215-217, where there are even more elaborations. 

79 Cf. Prior Analytics 1, 13, 32a, 18-20. Ijl, below, n. 84, quotes the definition of possible in 
his version of the argument. 

*"K. Usulal-Din, p. 85. 

S 'K. al-lrshad, pp. 53-54, 57; K. al-Shamil, pp. 352-382, with an elaborate discussion of the 
eight "principles" underlying the argument, and with an excursus showing that Mu'tazilites, who 
recognize the efficacy of human will, are guilty of inconsistency in employing the argument. 

82 Frankl, Ein Mu'tazililischer Kalam, p. 27. 

"K. Nihaya al-Iqdam, ed. A. Guillaume (Oxford and London, 1934), pp. 91-92; 94-96. 
""Mawaqif, VIII, p. 42. 

"TaftazanI, A Commentary on the Creed of Islam, trans. E. Elder (New York, 1950), p. 37. 
h6 al-Hidaya (Hobot ha-Lebabot), I, 7 (end). 
" Eshkol ha-Kofer (Eupatoria, 1836), §26. 



Particularization Arguments 



169 



Averroes and Maimonides reject the argument on the grounds that two deities 
might exist who have separate responsibilities, each creating a segment of the 
world and not interfering with the other. 88 Amidl rejects the argument because 
he cannot accept the response made by 'Abd al-Jabbar to the objection that two 
deities might always agree. As far as the argument from mutual interference goes, 
Amidl surrejoins, two deities might indeed exist who always agree; for the 
assumption of two deities does not entail even the possibility of conflict. The fact 
that when the two arc considered in isolation from each other, 89 one of them 
would be able to will an effect and his compeer would be able to will the contrary 
docs not mean that they would have the possibility of doing so at the same time. 
The situation, Amidl proceeds, may be clarified by considering a single divine 
agent. A single divine agent can, at any moment, will an event or its contrary, 
but he cannot conceivably will both the event and its contrary at the same moment. 
Similarly, each of two divine agents would be able to will an effect or its contrary 
when each and his exercise of will is taken in isolation; but it docs not follow 
that one of the two could will the effect while his compeer was willing the 
contrary. Two deities might then exist who do not have even the possibility of 
willing contrary events; and the argument from mutual interference collapses. 90 

Further arguments for the unity of the creator were current in Kalam circles. 
In (b) an extension or attenuation of the argument from mutual interference, it 
was contended that the hypothesis of two divine creators embodies a contradic- 
tion. Joseph ibn Saddiq explains: Should two agents produce the world cooper- 
atively, they would "stand in need of each other"; they would be weak; they 
would not be "eternal, inasmuch as the eternal cannot be weak"; and yet the 
world has been shown to have an eternal cause. "If, by contrast, one of the two 
agents produced the world by himself, ... the other would be weak and not . . . 
eternal." The existence of two divine creators is therefore impossible. 91 IjT puts 
the thought differently: If two deities were equally "powerful" (qddir), any pos- 
sible "object of activity" (maqdiir) for them would have an "identical . . . relation 
. . . to each"; and as a consequence, they could perform no action whatsoever. 
No action could be performed by them cooperatively; for a given single action 
is, by definition, performed by one and not by two agents. 92 Nor could an action 
be performed by one to the exclusion of the other. For since every object of 



Averroes. K. al-Kashf, ed. M. Mueller (Munich. 1859). p. 49; German translation, with 
pagination of Arabic indicated: Philosophic nnd Theologie von Averroes aus dem Arahischen ueb- 
ersetit, trans. M. Mueller (Munich, 1875); Maimonides. Guide to the Perplexed, I. 75( 1 ). 

""Amidl apparently has in mind Juwaynl, K. al-lrshad, p. 54. 

m Ghdyti al-Maram (Cairo, 1971), pp. 151-152. 

'" ha-'Olam ha-Qatan, p. 50. Similar reasoning appears in al-Muqammis, in Vajda, "Lc Problcmc 
dc 1'Unite dc Dieu." pp. 57-58; and Bahya, al-Hidaya [Hobot Im-Lelmbot), I. 7(7). Cf. also Shah- 
raMfmi, K. Nilutyu al-lt/dam, p. 92; Maimonides. Guide. I. 75(5). 

" 2 I understand that ijl is defining a single action as the action done by a single agent. Should two 
agents perform an action, we would in fact have two actions. 



170 



Part icularizat ion Arguments 



activity would have an identical relation to each of the two agents, neither of 
them could, to the exclusion of the other, undertake the activity — unless some 
wholly inexplicable factor were to materialize and "tip the scales," thereby deter- 
mining which of the two equally powerful agents should act. Inasmuch as equally 
powerful divine agents would be incapable of performing any action whatsoever, 
the world must be the handiwork of a single powerful divine agent. 93 

Saadia and Bahya advance (c) an argument that is recorded as well by Juwayni, 
where it is attributed to the Mu'tazilites and "many of our [i.e., Ash'aritc] 
scholars." 94 It goes: Creation requires that a creator be posited; the requirement 
is satisfied as soon as a single creator is acknowledged; hence a single creator is 
all that should be posited. 95 A specialized version is offered by ShahrastanI and 
recorded by Amidl. As was seen, one line of reasoning in support of the inference 
of a creator from creation had been that an agent would be needed to "tip the 
scales" in favor of the world's coming into existence. The act of tipping the 
scales, ShahrastanI, now contends, "points" to an agent tipping the scales, but 
not to "two agents." Consequently only a single agent who tipped the scales in 
favor of the world's coming into existence should be posited. 96 Juwayni, who 
refers to the general form of the present argument, and Amidl, who refers to the 
specialized form, reject the argument. They remark, cogently, that the "nonex- 
istence of a proof" is not tantamount to "proving the nonexistence of a thing." 
Accordingly, the failure of the proof from creation to demonstrate more than a 
single creator is nowise tantamount to its demonstrating that no more than a 
single creator exists. 97 Maimonides was familiar with all sides of the discussion. 
He records the Kalam argument to the effect that no more than one divine creator 
should be posited because no more than one had been demonstrated. He repeats 
the objection raised by Juwayni and Amidl. He even reports a rejoinder to the 
objection, and submits a surrejoinder of his own. 98 

The foregoing have been arguments for the unity of God the burden of which 
is, in each instance, that the world could not have more than one creator. In 
addition, Kalam writers put forward considerations that do not build on the prior 
inference of a creator from creation. A number of writers maintain (d) that only 
one deity can existence because two entities possessing the nature of the deity 
could not be differentiated. The explanation takes divers forms. 99 A common 
form (d.i), which has something of an Aristotelian cast, 100 goes: The deity is 



gl ljl. Mawiiqif, VIII, p. 41. Similar arguments appear in Juwayni, K. al-Shamil, p. 384 (attributed 
lo the Mu'ta/.ililes); and Amidi, Ghaya al-Maram, p. 154. 
94 Juwayni, K. al-Shamil, p. 387. 

95 Saadia, K . al-Amanat, p. 80; Bahya. al-Hidaya (Hobot ha-Lebabot), I, 7(3). 
96 K. Nihaya al-Iqdam, p. 93, intertwined with other considerations. 
97 Juwayni, K. al-Shamil, p. 387; Amidl, Ghaya al-Maram, p. 153. 
'"'Guide, I, 75(4). 

"The motif appears as well in John of Damascus, De Fide Orihodoxa, I, 5. 
""Metaphysics XII, 8, 1074a, 33-37. 



Particularization Arguments 



171 



incorporeal. Inasmuch as two identical incorporeal beings could neither occupy 
different places nor exist at different times, they could in no way be distinguished 
from each other. Consequently, no more than one entity possessing the nature of 
the deity can exist. 101 An alternate form (d.ii) has a purely Kalam cast. Two 
entities with the nature of the deity, it is argued here, could be differentiated from 
each other only if one or both contained an added element setting it apart. The 
added element would have the status of an accident; everything affected by an 
accident is known— through the proof for creation from accidents 102 — to belong 
lo the class of beings that come into existence; and the ultimate cause of the 
world cannot have come into existence. Hence, no more than one entity pos- 
sessing the nature of the deity can exist. 103 A closely related form (d.iii) was: If 
two entities possessing the nature of the deity were to be differentiated, one or 
both would have to contain an added clement setting it apart. The entity con- 
taining the added clement would be composite; what is composite is generated; 104 
what is generated cannot be the eternal first cause. Only one entity with the nature 
of the deity can, therefore, exist. 105 Still another closely related form (d.iv) 
incorporates concepts from Aviccnna's philosophy and runs: If two deities could 
be differentiated, one or both would, as before, have to be composite. But the 
cause of the universe is known to be necessarily existent by virtue of itself, and 
what is necessarily existent by virtue of itself cannot be composite. 106 Conse- 
quently, two deities could not be distinguished, and no more than one can exist. 107 
Several writers, finally, utilize the telcogical mode of thought and maintain (c) 
that the unity of the plan of creation discloses the unity of the creator responsible 
for the plan. 1 " 8 

Such were the Kalam arguments for the unity of God. A separate repertoire 
of dialectical arguments was utilized for establishing that God is incorporeal. 

(a) David al-Muqammis, Maturidl, BaqillanI, 'Abd al-Jabbar, Juwayni, Gha- 
zali, Joseph al-BasIr, Jcshua b. Judah, Judah Hallcvi, Fakhr al-DIn al-RazI, and 
Ijl, give the following argument: The Kalam proofs of creation have established 
that atoms, accidents, and bodies all come into existence. But the first cause of 



""Juwayni. K. al-Sluimil, pp. 384-385 (attributed to Mu'la/.ilitcs); Ghazali. al-lqtisdd. pp. 74- 
75; Joseph al-Basir, in Frankl, Em Mu'laziliiischer Kalam, p. 27; Shahraslani. K. Nihaya al-Iqdam, 
p. 93; Ibn Ezra, in Kerem Hemed, IV, p. 4; Maimonides, Guide, I, 75(2); Amidi. Ghaya al-Maram. 
p. 153 (a peculiar version). 

102 Above, p. 134. 

""Bahya, al-Hidaya (Hobot ha-Lebabot), I 7(6) (with a slight difference); lbn Ezra, in Kerem 
Hemed, IV, p. 4; al-Muqammis, in Vajda, "Le Problemc dc I'Unite de Dieu," p. 53. 
""Above, p. 147. 

" l5 Ibn Hazm, K. al-Fasl ft ul-Mital. . pp. 36-37; Bahya, al-Hidaya (Hobot ha-Lebabot), I. 7(4). 

'""Below, p. 296. 

I07 lji. Mawaqif, VIII, p. 39. 

K,8 Bclow, pp. 218,221,224, 230. 



172 



Part icularizat ion Arguments 



the universe is known to be eternal. 09 The first cause of the universe cannot, 
therefore, be a body or a constituent of a body. ' 10 

(b) An argument that enjoyed considerable currency went: Every body has 
particular dimensions to the exclusion of other dimensions that it might have; 1 1 1 
every body occupies a particular place to the exclusion of other places it might 
occupy; 1 12 and every body possesses particular qualities to the exclusion of other 
qualities it might equally possess." 3 A "particularizing agent" must assign the 
body its dimensions, its place, and its qualities. Inasmuch as every body is thus 
dependent on an agent outside it, the ultimate cause, which is dependent on 
nothing whatsoever, cannot be a body. 

(c) A number of writers laid down the premise that bodies arc incapable of 
creating other bodies; their justification of the premise was cither that induction 
reveals bodies to be incapable of creating other bodies;" 4 or that nothing can, 
in principle, create its like." 5 If a body cannot create another body, the being 
that created the physical world obviously cannot be a body. 

(d) Finally, Fakhr al-DIn al-RazI and Ijl borrow again from Avicenna's philos- 
ophy and offer a proof of the incorporeality of God which does not rest on the 
premise of creation. They argue that every body is composite; the ultimate cause 
of the universe is known by demonstration to be necessarily existent by virtue of 
itself; and what is necessarily existent by virtue of itself cannot be composite. 
The ultimate cause of the universe therefore cannot be a body. 116 

A first cause of the physical universe has now been reached which is eternal, 
which is not a mere nature or necessary cause, and which is one and incorporeal. 
The Kalam writers go on to argue that God possesses certain key attributes, 
attributes that the Ash'arite school construed as, in some sense, real distinct 
things within the deity, but that the Mu'tazilite school construed as having no 
real distinct existence. 

109 Above, p. 165. 

"°Vajda, "Lc Problemc de l'Unite dc Dieu," p. 50; Maturidi, K. al-Tawhid, p. 38; Baqillani, K. 
al-Tamhid, p. 25; 'Abd al-Jabbar, Sharh al-Usul, p. 218; Juwaynl, K. al-lrshad, p. 43; idem, K. al- 
Shamil, p. 411; Ghazali, al-Iqtisad, pp. 38-40; idem, al-Risala al-Qudsiya, Arabic, p. 18, English, 
p. 36; Joseph al-BasIr, in Frankl, Ein Mu ' tazililscher Kalam, p. 25; Schreiner, Studien ueber Jeschu 'a 
ben Jehuda, p. 39; Judah Hallevi, Kuzari, V, 18(6); Rail, K. al-Arba'in, p. 104; Ijl, Mawaqif, pp. 
21, 26. 

'"Baghdad!, K. Usui at-Dln, pp. 73, 77; Juwaynl, K. al-Shdmil, p. 412; ShahrastanI, K. Nihaya 
al-Iqdam, pp. 105-106; Ghazali, al-Iqtisad, p. 39; Ijl, Mawaqif, VIII, p. 26; AmidI, Ghaya al- 
Maram, p. 181 (with critique); Maimonides, Guide, I, 76(3). 

" 2 Juwayni, K. al-Shamil, p. 413; Jeshua b. Judah, in Schreiner, Studien ueber Jeschu' a ben 
Jehuda, p. 40(3); Ijl, Mawaqif, VIII, p. 20. 

" 3 Razi, K. al-Arba-in, pp. 104-105 (4;7); Ijl, Mawaqif, VIII, p. 26. 

" 4 'Abd al-Jabbar, Sharh al-Usul, pp. 221-222; Joseph al-Basrr and Jeshua b. Judah, in Schreiner, 
Studien ueber Jeschu' a ben Jehuda, p. 39(2) and n. 3. The argument is rather complicated. 

"'Juwaynl, K. al-Shdmil, p. 413; Ibn Saddiq, ha-'Olam ha-Qatan, p. 51. Also see Saadia. K. 
al-Amanat, II, introd., p. 78; AmidI, Ghaya al-Maram, pp. 184-185. 

" 6 RazI, K. al-Arba'in, p. 104(1); Ijl, Mawaqif, VIII, p. 21. See a similar argument for the unity 
of God, above, p. 171. 



Part icularizat ion Arguments 



173 



To establish that God is "powerful," 117 it was argued that only a powerful agent 
would be able to create a world;"" or, more specifically, that only a powerful 
agent would be to create a world as well-designed as ours discloses itself to be. 1 19 

To establish that God is "knowing," Kalam writers submit that an agent exer- 
cising power must know what he is doing, and hence a powerful agent possesses 
knowledge; 120 or else that the creation of a well-designed world, such as ours is 
discovered to be, implies knowledge. 121 

To prove that God is "alive," they reason that whatever has power together 
with knowledge must likewise have life; 122 or that only something alive would 
have the ability to create. 123 

To show that God possesses will, they argue that power entails will, and 
therefore every being possessed of power is possessed of will. 124 Or they contend 
that the decision to bring a world into existence after none had existed before 
involves the exercise of will. 125 Or else they argue more specifically that the 
decision to create the world at a particular moment or with a particular set of 
characteristics to the exclusion of equally possible moments and equally possible 
characteristics involves will. 126 Of all the divine attributes the present attribute, 



1 "See above, pp. 167, 169, where some writers connect power to eternity. 

""Ash'ari, K. al-Luma', §14; Maturidi, K. al-Tawhid. p. 45; Baqillani, K. al-Tamhid, p. 24; 
'Abd al-Jabbar, Sharh al-Usul, p. 151; al-Majmu'. p. 103; Juwayni, K. al-lrshad. pp. 61-62; K. al- 
Shamil, p. 621; Ghazali, al-Iqtisad. p. 81; Ibn Saddiq, ha-'Olam ha-Qatan. p. 57; Shahrastani, A". 
Nihaya al-Iqdam, p. 170; Razi, K. al-Arba'in, p. 129 (power excludes necessity); AmidI, Ghaya al- 
Maram. p. 45 (with reservations); Ijl, Mawaqif, VIII, p. 49 (power excludes necessity). 

""Ghazali, al-Iqtisad, p. 80; idem. al-Risala al-Qudsiya. Arabic, p. 20, English p. 40. 

l2 "'Abd al-Jabbar, K. al-Majmu', p. 107; Ijl, Mawaqif, VIII. pp. 66-67. 

m Ash'ari, K. al-Luma', §13; Maturidi, K. al-Tawhid, p. 45; Baqillani, A'. al-Tamhid. p. 26; 
'Abd al-Jabbar, Sharh al-Usul, p. 156; al-Majmu', p. 113; Juwaynl, K. al-lrshad, p. 61; idem, K. 
al-Shamil, p. 621; Ghazali, al-Iqtisad, pp. 99-100; idem, al-Risala al-Qudsiya, Arabic, p. 20, 
English, p. 41; Judah Hallevi, Kuzari, V, 18(7); Ibn Saddiq, ha-'Olam ha-Qatan, p. 57; ShahrastanI, 
K. Nihaya al-Iqdam, p. 171; Razi, K. al-Arba'in, p. 133; AmidI, Ghaya al-Maram, p. 45 (with 
reservations); Ijl, Mawaqif, VIII, p. 65. 

'"Baqillani, K. al-Tamhid, p. 26; 'Abd al-Jabbar, Sharh al-Usul, p. 161; Juwaynl, A'. al-Shdmil, 
p. 622; Ghazali, al-Iqtisad, pp. 100-101 ; idem, al-Risala al-Qudsiya. Arabic, p. 21, English, p. 41; 
ShahrastanI, A'. Nihaya al-Iqdam, p. 171 (not quite clear); Razi. A'. al-Arba'in, pp. 154-155; Ijl, 
Mawaqif. VIII, p. 80. 

'"Ash'ari, K. al-Luma'. §14; Juwayni, K. al-Shamil, p. 622. 

,24 'Abd al-Jabbar, Sharh al-Usul, pp. 107; 147; cf. Judah Hallevi, Kuzari, II, 6. 

'"Maturidi, K. al-Tawhid, p. 45; 'Abd al-Jabbar, Sharh al-Usul. p. 120; Razi, A'. al-Arba'in, p. 
1 29; and see above, p. 166, the argument that the cause of the world cannot be a nature or necessary 
cause. 

l26 BaqillanI, A'. al-Tamhid, p. 27 (in connection with the decision to create all things, and not 
just the world); Juwaynl, K. al-lrshad, p. 64; Ghazali, al-Iqtisad. p. 101 (not specifically in regard 
to the time the world as a whole came into existence); idem, al-Risala al-Qudsiya. Arabic, p. 21, 
English p. 41; Judah Hallevi. Kuzari. V, 18(9); Shahrastani, A". Nihaya al-Iqdam, p. 171; Razi, A". 
al-Arba'in. p. 147; AmidI, Ghaya al-Maram, p. 45 (with reservations); Iji, Mawaqif. VIII, p. 82. 
The contention survives in Muhammad 'Abduh, K. al-Tawhid (Cairo, 1966), p. 33; English trans- 
lation: The Theology of Unity, trans. I. Musa'ad and K. Cragg (London, 1966), p. 50. 



174 



Particularization Arguments 



will, is what distinguishes the Kalam concept of the deity from, for example, the 
Aristotelian concept. The intrinsic status of will is underlined by Ghazali, who 
insists that the existence of God has not been demonstrated until the cause of the 
universe is shown to possess will; 127 and it is likewise made explicit by those 
who follow their proof of a cause of the world with an argument showing that 
the cause of the universe could not possibly be a nature or necessary cause. 128 
But the centrality of will for the Kalam concept of the deity is undoubtedly 
recognized as well by adherents of the Kalam who do not dwell upon the point 
and who treat will routinely together with the other divine attributes. 

The upshot of Kalam natural theology, in fine, is that a first cause exists who 
brought the world into existence; that the cause is eternal, one, and incorporeal; 
that he is possessed of power, knowledge, life, and will. 129 Precisely which 
attributes are, and which are not, integral to the concept of the deity — and hence 
precisely where the proof of the existence of God ends and the arguments estab- 
lishing ancillary attributes begin — is, as has been mentioned, difficult to deter- 
mine. Yet it would seem that every adherent of the Kalam expected a proof of 
the existence of God to establish, at the minimum, the existence of an eternal, 
powerful, first cause of the universe. 130 At least some Kalam thinkers plainly 
regard will as integral to the concept of the deity. And some, perhaps, regard 
unity too as integral to the concept. 131 

2. Arguments from the concept of particularization 

At the heart of the arguments to be examined in the present section is the notion < 
that when an object has a given characteristic but could have alternative char- a 
acteristics, something must particularize the object in its characteristic or — the , 
phraseology can also go- — particularize the characteristic for the object. Some- » 
thing, that is to say, must choose the particular characteristic the object does have , 
from among the totality of possible characteristics it might have. The distinc- / 
tiveness of the particularization mode of thought can be brought out by contrasting 
it with the Aristotelian mode. The former, in its classic versions, supposes that 
all characteristics of physical objects are equally in need of an explanation; and 
the explanation provided is that each characteristic is the outcome of an arbitrary 
choice. The latter considers only certain characteristics of objects to be in need 
of an explanation; and its explanations are formulated in accordance with what 
it understands to be the laws of nature. To take an example, the particularization 
approach is as eager to ask why an object should be at rest and not in motion as 



'"Above, p. 3. 
128 Above, p. 166. 

'"Other divine attributes, of a theological character, are also established by the Kalam authorities, 
e.g., that God is "hearing,'' "seeing," and "speaking." 
l30 See above, p. 167. 



Particularization Arguments 



175 



to ask why an object should be in motion and not at rest. On the Aristotelian 
approach only the motion, and not the rest, of a physical object has to be explained, 
for empirical and analytic reasons — just as in a later physics, the single question 
requiring an answer is why an object has undergone a change of state; and the 
Aristotelian explanation of an object's motion is made in conformity with uniform 
laws of nature. Light is likewise cast on the particularization approach to the 
world by contrasting it with the tcleological approach. Whereas the teleological 
approach seeks out characteristics in objects which are so well designed that no 
human observer can fail to detect the imprint of a designing agent who chose 
intelligently, the particularization approach views the characteristics of objects A 
indifferently, and the choice it detects is an arbitrary one. x 

Arguments from the concept of particularization are usually associated with A 
an occasionalistic picture of the universe. When Kalam thinkers ask why a given / 
individual object came into existence at a given time or why the object has given 
characteristics to the exclusion of others, they ignore the commonsense answer 
that the time when the object came into existence or the characteristics possessed 
by the object are determined by the situation obtaining within the world prior to 
the object's coming into existence and by circumstances surrounding the object. 
The association of the particularization concept with an occasionalistic picture > 
of the universe is not, however, absolute. The concept could be applied as well / 
to general features of the world which would not be amenable to any natural ( 
explanation. General features that might be other than they are and that do not 1 
permit a natural or rational explanation must — it could be argued — surely reflect , 
an act of deliberate choice. 

The particularization notion has been traced to a detail of Stoic speculation. 
Chrysippus and certain adversaries debated whether human actions arc wholly 
determined by natural causes or whether, on the contrary, man has an autonomous 
power to opt for one course of behavior over another. The adversaries held that 
the human mind does have the ability to "incline" in favor of one course of 
behavior as against another, an ability "especially manifest" in instances where 
man is faced with "indistinguishable alternatives." Chrysippus responded that 
neither the "fall" of the "dice," nor the "inclination" of the "scales," nor any 
other event in the universe ever occurs spontaneously and "without a cause." ' 12 
The remark of Chrysippus regarding an inclination of the scales of an actual 
physical weighing apparatus may prefigure the metaphor of tipping the scales in 
Islamic literature. And the notion of tipping the scales is akin to the notion of i 
particularization; to say that the scales were tipped in favor of a given character- j 
istic is equivalent to saying that the characteristic was particularized for its subject. 

l32 Plutarch, Moralia: De Sioricorum Repugnantiis, 23; S. Horovitz, "Ucber den Einlluss des 
Stoicismus auf die Entwickclung der Philosophic bei den Arabcrn," Zeitschrift der Deutschen Mor- 
genlaendischen Gesellschaft, LV1I (1903), 190. See also N. Reseller, "Choice without Preference," 
Kam-Studien, LI (1959-1960), 143-166. 



176 



Particularization Arguments 



A nearer and more likely route through which the particularization notion may 
have been introduced into Arabic thought was the thrust and parry in connection 
with a recurring argument for eternity. The argument in question reasoned that 
creation is impossible because no moment in empty infinite time could have lent 
itself to the world's coming into existence in preference to the identical earlier 
and later moments when the world might have come into existence. A recurring 
response going back to Augustine and Philoponus was that God, through his will, 
would be capable of arbitrarily selecting one moment for creation in preference 
to the other equivalent moments at which the world might have emerged. 133 Here 
we have a given moment's being selected from among equally possible moments 
through a sheer act of will; and this, if not the sole source of inspiration for 
Islamic particularization arguments, undoubtedly fostered them. 

The particularization style of thought was long-lived. A vestige survives in 
Leibniz's "great principle . . . that nothing happens without a sufficient reason," 
a principle that compels one to ask "why there is something rather than nothing" 
and why, moreover, things "exist so and not otherwise." 134 Leibniz's principle 
of sufficient reason diverges, though, from the classic particularization principle 
in that the former is understood by Leibniz to operate with the highest rationality 
and not arbitrarily. 135 The philosophic problem of how a particular moment might 
have been selected for creation remained a topic for discussion in even later 
centuries. Hume and Kant 136 can still be discovered wrestling with it. 

The concept of particularization found many applications in Arabic philosophic 
literature. As has already been shown, it furnished Kalam thinkers with a rationale 
for defending their inference of a creator from creation. In addition, it supplied 
the nerve of a new proof of the existence of God when Kalam thinkers contended 
that, quite apart from creation, the characteristics exhibited by the world must 
have been selected by an agent outside the world. It supplied the nerve of a new 
method of proving the creation of the world when adherents of the Kalam and 
one nonadherent contended that the selecting out of particular characteristics for 
the world could not have been effected from eternity. And it served as a dialectical 
tool in miscellaneous contexts both connected with, and independent of, the issues 
of creation and the existence of God. 

Particularization arguments, it was seen in the previous section, constituted 
one of two general lines of reasoning whereby the inference of a creator from 
creation could be supported. Within that single line of reasoning, three or four 

'"Above, pp. 68-69. 

l34 Leibniz, Principes de la Nature el de la Grace, §7; cf. Theodicee, §14. 

133 Cf. A Collection of Papers which Passed between the Late Learned Mr. Leibnitz and Dr. Clarke 
(London, 1717), IV, §1: "In things absolutely indifferent there is no choice and consequently no 
election or will, since choice must have some reason or principle." 

136 Hume, Treatise of Human Nature, I, iii, 3; Kant, Critique of Pure Reason, A427/B455 (antith- 
esis of first antinomy). 



Particularization Arguments 



111 



subordinate strands can be distinguished. Three subordinate strands were rec- 
ognized by Juwayni, who makes the following perceptive observation: In "estab- 
lishing the need of what comes into existence for an agent bringing it into exis- 
tence," Kalam thinkers employed varying "formulations" and "terminologies." 
All the formulations presume that when an object contains a "possible charac- 
teristic," that is-to say, a given characteristic to the exclusion of others it might 
possibly have, the characteristic it docs have "depends on a particularizing agent." 
But each formulation focuses on a different type of characteristic. Attention may 
be directed, first, to "existence and nonexistence"; or secondly, to "the earlier 
appearance of some objects coming into existence and the later appearance of 
others"; or thirdly, to "the particularization of bodies in [or:by]" sundry "attri- 
butes, . . . shapes, or manners of composition," 137 In the first instance, the con- 
tention would be that what comes into existence might equally have remained 
nonexistent, and a particularizing agent must have selected existence for it in 
preference to nonexistence. In the second instance, the contention would be that 
what comes into existence at a given moment might have come into existence at 
an earlier or later moment, and a particularizing agent must have chosen the 
moment for the object to come into existence. In the third, the contention would 
be that a physical object might have come into existence with other attributes, 
another shape, and another composition than those it has; and a particularizing 
agent must have selected the features with which the object did come into exis- 
tence. These three strands of argumentation distinguished by Juwayni increase 
to four if we further differentiate between the contention which was put forward 
to the effect that a particularizing agent must have selected a time for the emer- 
gence of the world as a whole, and the contention which was put forward to the 
effect that a particularizing agent must select a time for the emergence of every 
object within the world. 

Four separate strands may, then, be differentiated in the deployment of the 
argument from particularization to support the inference of a creator from crea- 
tion. Three of the four strands have already been examined. Juwayni and others, 
it was seen, brought the particularization line of argumentation to bear on the 
selecting out of existence for the world in preference to nonexistence; they inferred 
the existence of the creator as the particularizating agent or the agent tipping the 
scales who made the selection. 138 Maturidl, without use of the term particular- 
ization, and Juwayni, who did use the term, brought the particularization line of 
argumentation to bear on the time when the world as a whole came into existence; 
building on their prior proof of the creation of the world, they inferred the 
existence of a creator as the agent who chose a particular moment for the world's 
appearance. 139 BaqillanI, without using the term particularization, and Baghdad!, 

'"K. al-Shamil, p. 272. 
138 Above, pp. 161-162. 
I3 " Above, pp. 160-161. 



( 



( 

^ 178 Particularization Arguments 

( who has the term, brought the particularization notion to bear not on the time 

when the world as a whole came into existence, but on the time when each of 
( the individual physical objects comprising the world came into existence. They 

built on their proof of the creation of every accident and body. And they presup- 
posed the Ash 'ante denial of natural causation. Within a physical universe where 
' natural causation reigns, the moment when individual objects emerge is deter- 

- mined by forces operating in accordance with natural law; but BaqillanI and 

Baghdad! supposed that the state of physical objects is nowise determined by 
( what takes place within the universe. They could therefore reason: Each object 

coming into existence might have come into existence at a different time from 
( the time when it did. Each object consequently had to have a time for its cmcr- 

( gence selected out by an agent apart from it. And since all bodies and accidents 

, are known to have been created, an agent must be posited who selected out the 

time for each to come into existence. 140 
( Side by side with his application of the particularization notion to the times 

when objects come into existence, BaqillanI — in a passage that has not been 
taken up yet — brings the notion to bear on the shapes and configurations pos- 
/ sessed by objects coming into existence. He is again presupposing that the state 

of a body is not determined by forces within the physical universe. He does now 
' expressly use the terms particular and particularizing ; and he writes: "Each body 

( in the world" has the "possibility (sihha) of receiving" a different "composition," 

that is to say, a different configuration of atoms from the configuration it has. 
^ What is "square" could be "round," and vice versa; and what has the "form of 

( one animal" could have "the form of another." Bodies, moreover, are constantly 

"transformed" from one "shape to another shape." Plainly, a body that "is par- 
' ticularized in [or:by] a specific, particular shape" cannot have been "particular- 

( ized in its shape by virtue of itself or merely by virtue of [its having] the possibility 

of receiving the shape." For then a body would "have to receive, at the same 
1 time, every shape it has the possibility of receiving," which of course does not, 

( and cannot, occur. The conclusion drawn by BaqillanI is that "whatever possesses 

a shape" received its shape through a "combining agent who combined it and an 
' intending agent who intended that it should be as it is." 141 

( Baqillanl's ostensible aim here is to support the inference of a creator from 

creation; he is ostensibly reasoning that since all bodies have come into existence, 
( and since they might have come into existence with a different set of charactcr- 

^ istics from the set they have, a particularizing agent must have brought them into 

existence with the characteristics they do have. The premise of creation is not, 
f however, required for the purpose. BaqillanI might have set the issue of creation 

C aside and still argued for a voluntary agent who assigns to bodies one out of all 

( 

140 Above, pp. 160-161. 

14 'AT. al-Tamhid, pp. 23-24. 



f 



V 



Particularization Arguments 



179 



the conceivable sets of characteristics they might have. He is, in other words, on 
the verge of a proof for a voluntary cause of the universe which dispenses with 
the premise of creation, a proof wherein a voluntary cause is inferred directly 
from the presence in things of particular characteristics. 

To repeat, Kalam writers adducing a particularization argument to support their 
inference of a creator from creation, contended either that the existence of the 
world would have to be selected in preference to the world's nonexistence; that 
a time would have to be selected for the coming into existence of the world as a 
whole; that a time would have to be selected for the coming into existence of 
every body in the world; or that a set of characteristics would have to be selected 
out for every body coming into existence within the world. 

Besides being deployed in support of the inference of a creator from creation, 
particularization arguments were utilized by the Kalam in other contexts. When 
establishing that the creator of the world cannot be a "necessary cause" or "nature," > 
Juwaynl argued that inasmuch as a necessary cause and a nature act in an unvar- 
ying manner as long as they exist, neither an eternal nor a nonetcrnal necessary 
cause or nature could be the cause of the world. An eternal necessary cause or 
nature could not be the cause of the world because an eternal necessary cause or 
nature would produce the world from eternity, whereas the world is known to 
have been created. But a noneternal necessary cause or nature — or, for that 
matter, any noneternal being — could also not be the cause of the world. For a 
nonetcrnal necessary cause or nature would need a "particularizing agent" to 
assign existence to it at the time when it came into existence; if that particularizing 
agent were noneternal, it would require another particularizing agent to assign 
existence to it; and so on ad infinitum. The creator therefore cannot, Juwaynl 
found, be anything except a "voluntary agent." 142 

Ijl established the unity of God by arguing that if two deities existed, neither 
could undertake any action unless a factor should inexplicably materialize and 
"tip the scales," thereby determining which of the two equally divine agents was 
to perform the action. 143 A number of Kalam writers proved the incorporeality 
of God by arguing that every body has particular dimensions, a particular loca- 
tion, and other particular qualities; every body accordingly depends on a "par- 
ticularizing agent" that assigns it its dimensions, place, and qualities; hence the 
first cause, being dependent on no agent, cannot be a body. 144 Kalam writers 
established the presence of will in the deity by arguing that the selection of a 
particular moment for the world to come into existence in preference to the infinite 
alternative moments when the world might have come into existence entails 



K. al-lrshad, pp. 28-29. 
'Above, pp. 169-170. 
'Above, p. 172. 
Above, p. 173. 



180 



Particularization Arguments 



The foregoing have been arguments pertaining to the issues of creation and 
the existence of God, but particularization arguments were resorted to elsewhere 
as well. They crop up constantly in the dialectical give and take typical of the 
Kalam, notably in the heavy dialectical web of real and artificial objections, 
rejoinders, surrejoinders, and sur-surrejoinders which later writers such as Juwaynl, 
RazI, and Amid! wove around every topic they touched. The language of parti- 
cularization could even be employed by philosophers of the Aristotelian school. 
They, however, use that language with no connotation of arbitrariness, but simply 
as synonymous with the language of natural causation. 146 

An application of the particularization concept which has not been mentioned 
yet but which is of interest for us is its use to establish the existence of accidents. 
The accident as an actual entity that carries a characteristic and, when conjoined 
to an inert atom or to a collection of atoms constituting a body, imparts the 
characteristic to them was an idiosyncratic feature of theoretical Kalam physics. 
The proposition that accidents do exist served as a key premise in the most 
distinctive Kalam proof of creation. 147 Yet the existence of accidents as actual 
entities is hardly obvious, and proofs had to be furnished. 

An argument for the existence of accidents which reportedly goes back to the 
ninth century ran as follows: If "motion," to take an example, were not an actual 
"thing" (ma'na) in the moving body, there could be no reason why one body is 
"in motion in preference to another" body, and no reason why a body "moves at 
the time it does move in preference to moving at an earlier time." Similarly, if 
"blackness" and "whiteness" were not "things" in bodies there would be no 
reason why a certain body is black or white "to the exclusion of another body's" 
being black or white. The circumstance that motion and rest, blackness and 
whiteness, and the "remaining" qualities, occur in some bodies but not in others, 
and the further circumstance that they occur sometimes but not at other times, 
demonstrates — the argument went — that they occur by virtue of "things," by the 
inherence in bodies of accidents construed as real entities. 148 

BaqillanI split the argument into two versions, the first of which considers why 
a body should be in motion at one time to the exclusion of another time, while 
the second considers why one body should be in motion to the exclusion of another 
body. The first version reads: A body is observed to be in "motion . . . subsequent 
to its being at rest" and to be at "rest subsequent to its moving." A body's being 



14 *See Alfarabi (?), Ta'liqat, in Rasa'il al-Farabi (Hyderabad, 1931), pp. 10, 14; Avicenna, 
Shifa': llahiyat, p. 411; De Anima, ed. F. Rahman (London, 1959), p. 229; Averroes, Tahaful al- 
Tahafut, ed. M. Bouyges (Beirut, 1930), p. 412; English translation, with pagination of the Arabic 
indicated: Averroes' Tahafut al-Tahafut, trans. S. van den Bergh (London, 1954); Maimonidcs, 
Guide, I, 73(10); 74(5). 

""Above, p. 137. 

u8 Ash'ari, Maqalat al-lslamiyln. ed. H. Ritter (Istanbul, 1929-1933), pp. 372-373, in name 
of Mu'ammar. Cf. Khayyat, K. al-lntisar (Beirut, 1957), §34; Wolfson, Philosophy of the Kalam 
(Cambridge, Mass., 1976), pp. 149 ff. 



Particularization Arguments 



181 



at rest or in motion must come about "either by virtue of [the body] itself or by 
virtue of a cause {'ilia). But if the body were in motion by virtue of itself, it 
could not |as long as it exists] ever possibly be at rest. The possibility of its being 
at rest after being in motion proves that it moves by virtue of a cause," by the 
presence in it of the accident of "motion." A similar analysis, BaqillanI adds, 
will cover "colors, tastes, odors, combination, . . . and the like," and show that 
each of them occurs through something actual, through an accident construed as 

I4Q 

an entity. 

In his second version of the argument for the existence of accidents, BaqillanI 
considers why one body should be in motion to the exclusion of another. A body 
in motion, he begins, must "move either by virtue of itself or by virtue of a thing 
(ma'na).'" But a body cannot "possibly be in motion by virtue of itself; for if 
such were the case, everything belonging to its genus" — everything of exactly 
the same nature — "which exists at the specific moment would have to be in 
motion. . . . The fact that members of the genus of mobile atoms and bodies are 
sometimes seen not to be in motion proves that atoms and bodies in motion do 
not move by virtue of themselves . . . but by virtue" of a thing, by the presence 
of the accident of "motion." 150 

The grounds adduced by BaqillanI for the existence of accidents are, it is to 
be noted, almost identical with the grounds he adduced for the existence of a 
particularizing agent who selects out the characteristics of bodies coming into 
existence. When inferring a particularizing agent, he reasoned that a body coming 
into existence with a certain shape or configuration cannot have been "particu- 
larized in its shape by virtue of itself or merely by virtue of [its having] the 
possibility of receiving the shape"; for then a body would "receive, at the same 
time, every shape it has the possibility of receiving." Inasmuch as shape and 
configuration cannot flow from the physical substratum of the body, the body 
must, he concluded, be particularized in its shape and configuration by a "com- 
bining . . . and intending agent" who determines in what form it should exist. 151 
When establishing the existence of accidents, BaqillanI reasons again that a body 
with a certain characteristic cannot have the characteristic "by virtue of itself"; 
for were such the case, the body would always have the characteristic, and every- 
thing capable of having the characteristic would also always have it. His conclu- 
sion here, however, is that inasmuch as the characteristic cannot flow from the 
physical substratum of the body, the body has the characteristic thanks to a thing, 
thanks to an accident construed as an actual entity. Virtually identical argumen- 
tation leads him in one context to an external agent who selects out characteristics 
for bodies coming into existence and in another context to the existence of acci- 
dents conceived as entities that inhere in bodies and impart characteristics to 
them. 

I49 A:. al-Tamhid, p. 18. 

"°lbid., p. 19. "'Above, p. 178. 



i 



182 Particularization Arguments 

, Other writers offer an argument for the existence of accidents which sounds 

like an argument for the existence of a particularizing agent. BaghdadI, who 

( employed a particularization argument to support his inference of a creator from 

/ creation, 152 points out in his discussion of accidents that "a body moves after 

being at rest" and that a body is "black after being white." The body, he proceeds, 

( surely does not move and is not black "by virtue of itself"; for the body "itself" 

is in existence both at the time when it does and the time when it does not move, 
both at the time when it is black and the time when it is not black. Inasmuch as 

1 motion and blackness cannot be due to the body itself, they must come about "by 

( virtue of a thing (ma'nd) that inheres in the body." 153 'Abd al-Jabbar, who did 

not employ a particularization argument to support the inference of a creator from 

( creation, writes in regard to accidents: A body acquires a certain characteristic — 

^ 'Abd al-Jabbar's example is the body's having its atoms in a state of "aggrega- 

tion" — when it could as well retain the characteristic it previously had. "The 

( situation is the same, and the conditions are the same. There is, hence, no 

. avoiding something, a particularizing factor (mukhassis), by the presence of 

which" it acquired the new characteristic; for "otherwise" the body "would not 

/ be [characterized] in the given fashion in preference to the contrary fashion." The 

Arabic term translated here as "particularizing factor" is the term I translated 

' earlier as "particularizing agent," the Arabic simply meaning "particularizer." 

( The partjcularizer that 'Abd al-Jabbar is positing in order to account for the 

. appearance of new characteristics in bodies is not an agent; it is "nothing . . . 

' other than the existence of a thing (ma'nd)" an actually existent accident. 154 

> Juwayni affirms the existence of accidents as actual entities on the grounds that 

an atom is observed to be "stationary" and subsequently is observed to be "in 
motion in a particular spot . . . distinct from the spot from which it started. . . . 

' Its being particularized in its spot is possible, not necessary, . . . and . . . stands 

. in need of a determining factor (muqtadin) that determines the positive particu- 

larization for it." The "determining factor" cannot be "the atom itself"; "for if 

( it were, the atom would have been particularized in the given spot as long as it, 

the atom, was in existence." The determining factor must be a "thing," which is 
"added to the atom," 155 that is to say, an accident construed as an actual entity. 

( BaqillanI thus employs virtually identical argumentation when establishing the 

existence of accidents as physical entities that are present in bodies and impart 

' characteristics to them, and when inferring the existence of an external agent 

( who selects out characteristics for bodies. BaghdadI and Juwayni offer an argu- 

ment for the existence of accidents which resembles their arguments for the 

' existence of a particularizing agent who selected a time for individual objects to 



'"Above, p. 161. l "K. Usui al-Dln, p. 37. 

15 *Sharh al-Usul, p. 96. The argument, to be precise, is proving the existence of the most basic 
and general sort of accidents, the akwan. 
l "K. al-lrshad, pp. 18-19. 



< 



Particularization Arguments 



183 



come into existence, or who selected existence for the world when it came into 
existence, or who selected one time for the world to come into existence in 
preference to alternative times. 156 'Abd al-Jabbar docs not use the particulari- 
zation mode of reasoning when inferring a creator from creation, but does have 
a particularization argument for the existence of accidents; and he goes as far as 
to call the accident a "particularizing factor" or, more literally, a "particularizer," 
the very term whereby others designated the particularizing agent. It is hard to 
avoid asking why both the creator as an external particularizing agent and the 
accident as an inhering particularizing factor are required. Why, to be precise, 
must an external agent be posited if accidents account for the characteristics of 
atoms and bodies, and why must accidents as actual inhering entities be posited 
if the external particularizing agent can account for those characteristics? 

An answer to the first half of the question is supplied by the Kalam proofs of 
the generation of accidents. 157 If the accidents that impart characteristics to atoms 
and bodies come into existence, an added external cause has to be posited who 
brings the accidents into existence. The step from the coming into existence of 
accidents to a cause bringing them into existence might either be treated as self- 
evident, supported by analogy, or supported through a particularization argu- 
ment. 158 Should a particularization argument be used, the reasoning could be 
either that a particularizing agent must select out existence for the accident, 
conceived as a particularizing factor, in preference to nonexistence; that a par- 
ticularizing agent must select out a moment for the accident to come into exis- 
tence; or that a particularizing agent must decide which accident is to be joined 
to which atom. 159 

An answer to the second half of the question is indicated by BaqillanI. Besides 
giving the two versions of the argument for the existence of accidents already 
quoted, BaqillanI advances a totally different argument. He now approaches the 
subject of accidents not from the side of characteristics observable in a body, but 
from the side of the external agent producing the characteristics. When an agent 
"exercises power," he asserts, the agent's power must be "attached to some object 
of power" which possesses actual existence, just as "knowledge ... is attached 
to an object of knowledge" which possesses actual existence, and "memory, . . . 
to an object of memory" which possesses actual existence. When, for instance, 
the agent "exercises his power ... in moving a body," the exercise of power 
must have an actually existing "object of power," it must give rise to some actually 
existing thing in the body moved. And the object of power, the actually existing 
thing produced by the agent, can be nothing other than the concrete accident of 
motion in the moving body. 160 BaqillanI has hereby explained in effect why the 
existence of accidents as actual entities has to be accepted in addition to the agent 



156 Above, pp. 161-162. '"Above, pp. 137-138. 

""Above, pp. 155, 159. 

'"Above, pp. 160-162, 178. lw K. al-Tamhid, p. 19. 



184 



Particularization Arguments 



who selects out characteristics for bodies: When an agent acts, his action must 
bring about something concrete in the body acted upon. 

Juwayni does not, as far as I could detect, expressly address the question why 
both a particularizing agent and an accident construed as a particularizing factor 
must be posited. But a subtle answer to both halves of the question can be 
extracted from his discussion of accidents. Juwayni was troubled by an objection 
to the theory of accidents which was of the 'third man' type. 161 An accident, the 
challenge went, is supposedly the factor through which an atom is, for example, 
"particularized" in a certain "spot." But if an atom occupying a given spot has 
to be particularized in that spot in preference to others, the accident responsible 
for particularizing it there also has to be "particularized" for the atom with which 
it is connected in preference to all the other atoms with which it might be con- 
nected. And if what particularizes the atom in the spot it occupies is a "deter- 
mining factor," to wit an accident, which has the status of a real thing, the accident 
too should have to be particularized for its atom by a further determining factor, 
which would likewise have the status of a "thing." The further determining factor 
particularizing the accident in its atom would, in turn, have to be selected out 
and particularized for its accident to the exclusion of other accidents by still 
another "thing"; and "an infinite regress would ensue." 162 

Mu'ammar, the earliest Kalam figure known to have formulated the argument 
for the existence of accidents from the concept of particularization, realized that 
the argument, if allowed to run its logical course, ends in an infinite regress. But 
Mu'ammar, instead of recoiling from paradoxes, could embrace them with relish; 
and an infinite regress of accidents did not discomfit him in the least. He is 
reported to have maintained that motion occurs in a body by virtue of a "thing" 
{ma'na), and that the thing — the accident of motion — is connected to the body 
by an infinite series of other "things," all of which come into existence "at the 
same moment." 163 Juwayni, by contrast, was incapable of admitting an infinite 
regress of accidents in every atom or body; and he faced the objection that if 
each atom receives its characteristics through an accident conceived as a parti- 
cularizing factor, the accident would have to be particularized for its atom by a 
further factor, which is particularized by still another, ad infinitum. 

The objection, writes Juwayni, can be handled in two ways. It can, in the first 
place, be sidestepped by understanding that each accident is unique and suitable 
only for its own atom. The accident would accordingly "be particularized in its 
subject by virtue of itself," and no further "thing" would be required in order to 
determine which accident is to inhere in which body. 164 In the second place, the 

,6l Cf. Plato, Parmenides, 134; W. D. Ross's edition of Aristotle's Metaphysics, Vol. I (Oxford, 
1924), p. 195. 

t62 K. al-Shamil, p. 174. 

'"Ash'ari, Maqalat al-lslamiyin, p. 372; Khayyat, K. al-Intisar, §34; cf. Wolfson, Philosophy 
of the Kalam, pp. 149-162. 

IM K. al-Shamil, p. 174. The determination of a particular time for the accident to appear is not 
explained. 



Particularization Arguments 



185 



objection can be resolved as follows: Each "accident" can be understood to be 
"particularized in its subject by the intention of an intending agent" who decided 
to "particularize it" specifically there. 165 The accident would thus not need a 
further immanent factor to particularize it in its atom, and no infinite regress 
ensues. This second way of explaining how a given accident comes to be con- 
nected to a given subject indirectly answers the question why accidents are insuf- 
ficient in themselves to account for the occurrence of characteristics in objects 
and why a particularizing agent must be posited as well. The answer indicated is 
not very different from the answer implied by the Kalam proof of the generation 
of accidents: In order to avoid an infinite regress, a particularizing agent must be 
posited who determines that a given accident should inhere in a given atom to 
the exclusion of the other atoms in which the accident might inhere. 

Juwayni has another point to make. The imaginary or real opponent who is 
challenging the theory of accidents might counter that the foregoing exposition 
permits accidents to be dispensed with altogether. If the principal accident can 
be assigned to an atom through an external agent with no help from further 
accidents, why, the opponent may riposte, might not a characteristic be assigned 
directly to its atom by the external agent with no help from any accident what- 
soever. Juwayni fends off the riposte. The real or imaginary opponent challenging 
the theory of accidents has, he avers, forgotten the diverse natures of atoms and 
accidents. Accidents, in the dominant Kalam physical scheme, remain in exis- 
tence for no more than a moment of atomic time, whereas atoms enjoy an exis- 
tence extending over a number of moments. 166 An agent, even the divine agent, 
can exercise his "power" solely at the moment when he brings the object of his 
power into existence. Since the divine particularizing agent brings accidents into 
existence anew at each moment, he can at each renewal assign a given accident 
to a given atom; and no supplementary factors are needed to link the accident to 
its atom. But the divine agent allows atoms to exist over a stretch of moments, 
and he could by his own act assign a characteristic to an atom — locating the 
atom, for example, in a particular spot — for no more than the first moment. An 
actual real accident must therefore constantly inhere in the atom — or, to be more 
precise, must constantly be recreated in the atom — in order to locate the atom in 
its spot for all subsequent moments. 167 Juwaynl's position is perhaps problemat- 
ical, 168 but it does, after a fashion, explain why the particularizing agent is 
insufficient and why concrete accidents have to be assumed as well. If the par- 
ticularizing agent acted directly on an atom, he could do so for the duration of 

165 K. al-Shamil, p. 175. Mu'ammar reportedly held that God does not create accidents; Ash'ari, 
Maqalat, p. 199. 

' 66 Cf. Wolfson, Philosophy of the Kalam, pp. 522 ff., where no less than eight theories of the 
duration of accidents are differentiated and discussed. 

167 K. al-Shamil, p. 175. Juwayni adds that since the accident is needed for subsequent moments, 
it would be needed for the first moment as well. 

I6 "lf an accident is required to locate an atom in a particular spot and the accident exists for only 
a moment, it is hard to sec how the atom can be described as existing for more than a moment. 



186 



Particularizalion Arguments 



no more than a moment, and there would be nothing to impart qualities to the 
atom during the remaining moments. 

We have this picture: BaqillanI, BaghdadI, ' Abd al-Jabbar, and Juwaynl, estab- 
lish the existence of accidents as real things through the consideration that some 
factor must be present in a body or atom to particularize the body or atom in 
each of its characteristics. In a separate context, BaqillanI supported the inference 
of a creator from creation through the consideration that a particularizing agent 
must select out the characteristics possessed by bodies coming into existence; 
BaqillanI and BaghdadI supported the inference through the consideration that a 
particularizing agent must select out the time when individual bodies come into 
existence; and Juwaynl supported the inference through the consideration that a 
particularizing agent must select out existence for the world in preference to 
nonexistence, and a moment for the world to come into existence in preference 
to alternative moments when the world might have come into existence. As to 
why both a particularizing agent and a particularizing factor are needed, half the 
question is answered by Kalam proofs of the generation of accidents. Granting 
the existence of accidents, a particularizing agent must be posited who selects 
out existence for the accident, or a time for the accident to come into existence, 
or an atom for the accident to inhere in. BaqillanI indicates an answer to the other 
half of the question. He explains why, assuming the existence of a particularizing 
agent, the existence of accidents as the carriers for characteristics has to be 
recognized: When an agent acts it must produce something concrete in the body 
acted upon. Juwaynl indicates answers to both halves of the question. The par- 
ticularizing agent is needed so that one accident to the exclusion of another can 
be assigned to a given atom without an infinite regress' ensuing; and the accident 
is needed so that an atom can retain characteristics, such as location in a particular 
spot, over a stretch of moments. The discussion, as cannot have been missed, is 
entirely rooted in the peculiar Kalam physical universe, with its inert atoms, 
accidents construed as actual entities, atomic time, and denial of natural causality. 

And the dialectic again approaches a proof for the existence of God in which 
the issue of creation is set aside. One of the contentions whereby BaqillanI 
supported the inference of a creator from creation was already seen to border on 
an argument for the existence of a particularizing agent who assigns character- 
istics to all bodies, whether or not the world has come into existence. 169 The 
dialectic of the theory of accidents has now maneuvered Juwaynl into stating that 
to avoid an infinite regress, a particularizing agent must be posited who assigns 
accidents to atoms. Having been led to that statement, Juwaynl might have based 
the existence of the cause of the universe on the need to posit a particularizing 
agent who assigns accidents to atoms and bodies whether or not atoms and bodies 
are created. As it turns out, Juwaynl and others do advance such an argument. 



Above, p. 178. 



Particularization Arguments 



187 



3. Particularization arguments for the existence of God without the 
premise of creation; particularization arguments for creation 

A proof of the existence of God from the concept of particularizalion without 
reference to creation is perhaps to be detected in the Ikhwan al-$afa\ the so- 
called "Brothers of Purity." The Ikhwan remark in one passage that a "body 
cannot move in every direction lor: to every spot] at the same time," and they 
conclude that a body's "movement in a certain direction to the exclusion of 
another must be due to a cause." 170 The meaning seems to be that a particularizing 
agent must exist who arbitrarily chooses the direction in which bodies and espe- 
cially — as the text indicates — the heavenly bodies move. The passage happens, 
however, to be interlaced with teleological motifs and may be animated solely 
by the teleological, and not the particularization, outlook. The Ikhwan may, in 
other words, be concerned only with evidence of design in the world, not with 
evidence of arbitrariness. The meaning of the passage would then be that since 
the movements of the celestial spheres in their several directions disclose design, 
they arc undoubtedly the work of a divine designer. 171 

A fully conscious argument for a voluntary particularizing cause is given in a 
composition of Juwaynl's from which I have not quoted yet. The composition, a 
later work, undertakes to establish the existence of God by "methods" Juwaynl 
had "hitherto not pursued," methods that he pronounces the "most useful and 
finest" he had ever met. 172 In the new procedure, Juwaynl notes that everything 
in the physical universe which is observable and, by analogy, everything not 
observable as well, has "possible" characteristics. Any given body might, for 
example, "conceivably" have a "different shape" from the one it has. "What is 
at rest" could "conceivably" be "in motion," and vice versa. Physical objects 
that move upwards might move downwards. Objects that perform a circular 
motion, such as the heavenly bodies, might move in other orbits. The stars might 
be arranged in the heavens differently. And the world as a whole might have an 
alternative location in space. Since the parts of the world and the world as a 
whole could be different from what they arc, "a determining agent" (muqtadt) m 
must have selected out for the parts of the world, and for the world as a whole, 
the characteristics that they do have. 174 Juwaynl goes on to show, with the aid of 
typical Kalam considerations, that an agent who selects out characteristics and 
assigns them to the world would possess the familiar properties of the deity, 
namely unity, incorporeality, power, knowledge, life, and will. 175 He thus arrives 



""Ikhwan al-Safa - , Rasa'il (Beirut, 1957), 111, p. 336. 
"'See below, p. 225. 

m al-'Aqida al-Nizamiya (Cairo, 1948), pp. 8, 13; German translation, with pagination of the 
Arabic indicated: Das Dogma der. imam al-Haramain al-Djuwayni, trans. H. Klopfer (Cairo, 1958). 
171 In other contexts 1 translated this term as "determining factor." 
ni al-'Aqida al-Nizamiya, pp. 11-12. 

"'ibid., pp. 17, 29; the arguments are similar to those examined above pp. 166 ff. 



188 



Particularization Arguments 



at a single incorporeal cause of the world, a cause possessed of power, knowledge, 
life, and will; and he has not had to start with creation. From the Kalam stand- 
point, his argument has the virtue of at once establishing a specifically voluntary 
cause of the world, it being evidence of the arbitrary exercise of will that leads 
him to a cause of the world. His argument likewise has virtues from a non-Kalam 
standpoint. It looks at the characteristics of things without insisting on the exis- 
tence of accidents as actual entities. It looks, moreover, not merely at individual 
characteristics of individual objects within the world, characteristics that com- 
monsense and Aristotelian philosophers ascribe to the workings of natural forces, 
but also to structural features of the world which cannot easily be traced to 
immanent natural forces. 

ShahrastanI cites the argument in Juwaynl's name, although the evidence of 
arbitrary choice in the world which he adduces diverges somewhat from the 
evidence offered by Juwaynl; and he agrees with Juwaynl's assessment of the 
argument, finding it to be a "superlatively fine and perfect . . . method" for 
proving the existence of God. Shahrastanl's version runs: The constituents of the 
physical universe, to wit "earth, . . . water, . . . air, . . . fire, . . . and ... the 
spheres," might possibly occupy alternative places, might have an alternative 
"shape and magnitude," and might be "larger or smaller" than they are. But 
"whatever is particularized in a certain way ... to the exclusion of other . . . 
equally possible ways" is judged "by the necessity of intellect." 176 to stand in 
"need of a particularizing agent." All the parts of the world depend, therefore, 
on a particularizing agent. 177 An agent capable of selecting between equal pos- 
sibilities plainly acts by "power and choice," not by "nature." The agent upon 
which the universe depends must accordingly be possessed of power and will. 178 
Later chapters of Shahrastanl's book explain why a cause of the world would 
possess the remaining divine attributes; and the explanation given there is intended 
to supplement all his arguments for a cause of the world, including the present 
argument, and raise them all to the level of complete proofs of the existence of 
God. 

Fakhr al-DIn al-RazI too has an argument for a particularizing agent who is 
the cause of the world, but Razl's argument returns to a wholly occasionalistic 
framework by focusing exclusively on the characteristics of individual bodies. 
"Bodies," he contends, "are similar in their quiddity and essence," and as a 
consequence, "any attribute" that is connected to a given body might equally be 
connected to "other bodies." Any body with a "particular attribute" hence "stands 
in need ... of a particularizing agent and an agent tipping the scales" who 
selects out the attribute for it. And since the world consists of bodies that might 



'See above, p. 161. 

K. Nihaya al-Iqdam, pp. 12, 14. 

'Ibid., p. 14. 



Particularization Arguments 



189 



have alternative attributes, the world in its totality stands in need of a particu- 
larizing agent. 179 Razl's formulation of the argument is copied by Ijl. 180 Both 
RazI and Ijl bracket it with other arguments for a cause of the world, and the 
considerations whereby they establish that the cause of the world is eternal, one, 
incorporeal, and so on, are designed to supplement all their arguments and render 
them all complete proofs of the existence of God. 

The difference between the standard Kalam procedure for proving the existence 
of God and the procedure just examined merits comment. The standard procedure 
began by laying down a set of premises, among which is the existence of acci- 
dents; and a particularization argument provided one rationale for affirming that 
accidents indeed exist. After the premises had been laid down, the conclusion 
was drawn that the world was created. Then a creator was inferred from creation; 
and a particularization argument could again be called upon to support the infer- 
ence. In contrast to the standard procedure, the new procedure is more direct. 
Without the customary premises, with no insistence on the existence of accidents, 
and without first establishing the creation of the world, JuwaynT and the others 
take a thread from the standard procedure, the thought that the world and its parts 
might have alternative characteristics and that something must arbitrarily select 
out the characteristics the world and its parts do have. A voluntary particularizing 
agent is inferred directly from the arbitrary selection of characteristics for the 
parts of the world and for the world as a whole. 

Together with its role in a new, more straightforward proof of the existence of 
a cause of the universe, the particularization concept supplied the nerve of a new 
argument for creation. 'Abd al-Jabbar and his anonymous editor record such an 
argument, having as its pivotal premise the assertion that nothing can "particu- 
larize" an eternal body "in one spot to the exclusion of another." Every body, 
the reasoning goes, obviously has the possibility of occupying alternative loca- 
tions in space. Since nothing might have particularized an eternal body in a single 
spot to the exclusion of others, an eternal body would have to occupy either 
"every spot" simultaneously or no "spot whatsoever." Both suggestions are pre- 
posterous. Hence no eternal body can exist; and the body of the universe must 
have been created. 181 The premise affirming that nothing could particularize an 
eternal body in one spot to the exclusion of another is not elucidated in 'Abd al- 
Jabbar's report, but Kalam literature reveals how it might have been defended. 
Several Kalam writers, as will appear, argue that the act of particularization 
involves will and that the exercise of will is incompatible with eternity. An 
underlying and unstated step in the argument recorded by 'Abd al-Jabbar may 
accordingly be that inasmuch as the particularization of a body in a specific spot 
would involve will, it cannot have occurred from eternity. 

" 9 K. at-Arba'm, p. 84; cf. Muhassal. pp. 107-108. 
K0 Mawaqif, VIII, p. 4. 
m K. al-Majmu', pp. 63-64. 



190 



Particularization Arguments 



An argument for creation from the concept of particularization is advanced by 
Juwayni in connection with the argument already quoted wherein he employed 
the particularization concept to establish a cause of the world. Juwaynl's object 
was to provide a combined proof of the existence of God and creation, although 
I have disentangled the components and discussed them separately. 182 In the 
passage that has been quoted, Juwayni arrived at a particularizing agent who 
arbitrarily selects a set of particular characteristics for the world and its parts. 
He thereupon proceeds: A "necessary cause" is incapable of selecting between 
equivalent possibilities, and if faced with equivalent possibilities must embrace 
them indiscriminately. For example, a purging medicine is incapable of working 
on the right side of the body to the exclusion of the left and perforce affects both 
sides equally. When an arbitrary choice has been made between equivalent pos- 
sibilities, a voluntary, and not a necessary, cause is therefore responsible. 183 Now 
whereas a necessary cause acts unvaryingly as long as it exists and, if eternal, 
would act eternally, a voluntary cause cannot act from eternity. And inasmuch 
as the arbitrary selection of characteristics for the world does show the world to 
be the handiwork of a voluntary cause, the world — whether only its form, or its 
matter as well, is not made explicit by Juwayni 184 — cannot have been brought 
into existence from eternity but must have been created. 185 The keystone of the 
argument, the rule that a voluntary cause cannot act from eternity, is not so much 
explained by Juwayni as postulated. An "object of will," he asserts, is something 
"particularized that did not exist and subsequently existed . . . while, on the 
contrary, the existence of what exists eternally cannot possibly be dependent on 
will. ... In general, . . . what exists through will is an effect produced by a 
voluntary agent who brings it about in conformity with his will, whereas what 
exists eternally is not effected [by an agent at all]." 186 Juwayni, it is to be noted, 
is saying two things here: that eternity is incompatible with the exercise of will, 
and also that eternity is incompatible with a cause's having brought an effect into 
existence. The two points will reappear in writers to be discussed presently; the 
second point, moreover, is a perennial motif with a long history in medieval 
philosophy. 187 

Fakhr al-DIn al-RazI has an argument from the particularization concept which, 
although ostensibly a proof merely of creation, could, like Juwaynl's argument, 
stand as a combined proof for both creation and the existence of a cause of the 



182 ShahrastanI, K. Nihaya al-Iqdam, p. 14, docs the same. 

183 I am using language that Juwayni would reject. He would insist that a being who acts voluntarily 
should be called an agent, not a cause. 

""Perhaps Juwayni was thinking that the particularization of atoms in a given spot entails their 
creation. 

m al-'Aqida al-Nizamiya, p. 12. 

,86 Ibid., p. 12. 

""See above, p. 3; below, pp. 210, 387. 



Particularization Arguments 



191 



world. Every body and the world as a whole, RazI begins, are of "finite mag- 
nitude," and what is of finite magnitude could "conceivably" be larger or smaller. 
A "particularizing agent" musl, as a consequence, have "lipped Ihe scales" and 
selected out the exact magnitude every body has from among the alternative 
magnitudes it might have. A particularizing agent selecting between equal alter- 
natives is a "voluntary agent," 188 an agent operating by "choice and intent." But 
the "intent to produce something" is operative during the period of the thing's 
"nonexistence" or, if one prefers, at the "moment of its coming into existence; 
and in either event, anything coming about through a voluntary agent" is preceded 
by nonexistence and "is created." Since every body and the world as a whole 
have their magnitudes assigned by a voluntary agent, the world and everything 
contained therein have been created. 189 An abbreviated version of the same train 
of thought is found in Amid! 190 and IjT. 191 

RazI has an additional argument for creation from the concept of particulari- 
zation, and there he proposes a separate reason for the rule that a particularizing 
agent must precede its effect. The reason in the argument of Razl's just examined 
was that will excludes eternity; the reason now is to be that the eternity of an 
object is excluded by its having been brought into existence. Both reasons were 
already to be discovered in Juwayni. 192 

The world, RazI submits, is "possibly existent" and what is possibly existent 
has need of an external agent that "tips the scales" in favor of its existence. 193 
To maintain that something needs an "agent" to tip the scales in favor of its 
existence when it already exists would be absurd. The world must consequently 
have had the scales tipped in favor of its existence when it did not yet exist; and 
the world was created. 194 

The contrast between these arguments for creation and the standard Kalam 
procedure for proving creation and the existence of God again merits comment. 
The standard procedure set forth an elaborate proof of creation. It thereupon 
inferred a creator, often on the grounds that a particularizing agent must have 
selected a moment for the world's emergence or other characteristics for the world 
and the parts of the world when they came into existence. And it added that a 
particularizing agent must possess will. The new arguments borrow a thread, the 
particularization line of reasoning, from the standard procedure; and they invert 
the sequence of thought, concluding, rather than commencing, with the creation 
of the world. They observe that sundry characteristics of the world disclose an 
arbitrary choice between equivalent alternatives. From the evidence of arbitrary 

m K. al-Arba'in, pp. 27-29. Very similar reasoning appears in Shahrastani, K . Nihaya al-Iqdam, 
p. 13. 

""Ibid., pp. 17, 29. 

™Ghaya al-Maram, pp. 250-251. 

"'Mawaqif. VII, p. 227. 

192 Above, p. 190. l93 Sce above, p. 162. 

'""K. al-Arba'in, pp. 30-31; see below, p. 387, n. 54. 



( 

192 Particularization Arguments 

choice they infer an agent who exercises will. And they conclude that an agent 
( exercising will cannot have acted from eternity but must have brought the world 

into existence after not having done so. 
Averroes was familiar with Juwaynl's argument for creation from the concept 
' of particularization; he records it with explicit reference to the work of Juwayni's 

( in which it appeared. The argument, Averroes finds, rests on "two premises." 

The first states that "the world and everything therein" might have different 
[ characteristics from those they have; and the second states that whatever might 

( have alternative characteristics must be brought into existence because it depends 

on an "agent," more. specifically a "voluntary . . . particularizing agent," who 
( "fashioned it in one possible mold in preference to another" 195 As might be 

f expected, Averroes admits neither premise. 

In connection with the first premise, he recognizes a distinction between the 
( individual characteristics of individual objects within the world and characteris- 

C tics touching the structure of the world. The supposition that the characteristics 

of individual objects might be other than they are is dismissed by him as "patently 
( false," on the grounds that such characteristics are, unquestionably, determined 

I by natural forces. As to characteristics touching the structure of the world, the 

possibility of their being other than they are might, Averroes concedes, seem 
( plausible; for the "causes" of general phenomena — as, for example, the move- 

; ments of one celestial sphere to the "west" rather than "east," and the remainder 

to the "east" rather than "west" — are often "hidden from man." Nevertheless, 
( the scientific presumption should in every instance be that structural features of 

the universe are unqualifiedly "necessary" or at least represent the "best" and 

"most perfect" adaptation of natural objects for the functions they fulfill. It can 
C thus not be taken for granted either that individual or general characteristics in 

the world might be different from what they are, and that certain characteristics 

have hence been selected out arbitrarily in preference to others. 196 
( That is Averroes' refutation of Juwaynl's first premise. Averroes likewise rejects 

the premise stating that what is dependent for its characteristics on a voluntary 

agent must have been brought into existence after not having existed. Juwaynl 
( had explained, or postulated, that an "object of will" cannot be eternal. Averroes 

rejoins that an eternal agent possessed of an eternal will not only would be able 

to exercise his will eternally, but could not help doing so. Consequently, Averroes 
I clinches his refutation, even if the world were to reveal evidence of arbitrary 

selection, the agent making the selection would be capable of acting eternally, 

and the world need not have been created. 197 
( Maimonides records a Kalam argument for creation from the concept of par- 

ticularization which echoes both Juwaynl's argument and Shahrastanl's refor- 

j ■ m K. al-Kashf, pp. 37-38, 40. 

1 1 '"ibid., p. 38. Averroes undoubtedly has in mind the passage of Ghazali, quoted below, p. 195. 

| " 7 Ibid., p. 40. Cf. above, p. 76. 



Particularization Arguments 



193 



mulation of it. 19S Like Averroes, Maimonides rejects the supposition that the 
characteristics of individual objects within the world might be other than they 
are; he too traces them to natural forces operating in conformity with natural 
law. 199 Maimonides docs not, however, rule out the pertinence of the particular- 
ization notion to structural features of the world. He leaves open the possibility 
that these might be other than they arc, and that a particularizing agent can be 
inferred who selected out structural features the world does have over features 
the world might have. In addition, Maimonides differs from Averroes on the 
subject of will; he agrees with Kalam thinkers that the exercise of will is incom- 
patible with an eternal product. Because he can accept much of the present 
argument, Maimonides deems it the "best" of the Kalam arguments for creation, 
and he subsequently will rework it in what he believes to be an unexceptionable 
form. 200 

Maimonides also records an argument for creation, attributed by him to one 
of the "later" Kalam thinkers and described by him as a variation of the "pre- 
ceding," which turns on the world's being "possibly existent." The argument is 
identical with one in Fakhr al-DIn al-RazI. 201 What is "possibly existent," the 
reasoning goes, "has the possibility of existing as well as . . . not existing" and 
must, if it does exist, have had the scales "tipped in favor" of its existence. But 
an agent tipping the scales in favor of existence would not have acted from 
eternity. Consequently, the world, which is possible existent, cannot have existed 
from eternity. This train of reasoning is entirely rejected by Maimonides. The 
term "possibly existent," he explains, has two connotations, being applicable to 
what is, as well as to what is not, eternal. 21 ' 2 To be possibly existent is therefore 
by no means tantamount to having the possibility of not existing. An object can 
be known to have a possibility of not existing only if it is known to be both 
possibly existent and noneternal, whereas anything possibly existent and eter- 
nal — as "our adversary," that is to say, Avicenna and his school, judged the world 
to be — would not have the possibility. 203 By taking for granted that everything 
possibly existent docs have the possibility of not existing and hence does need to 
have the scales tipped in favor of its existence, the argument under consideration 



,qs Guide, I, 74(5). Some of the features that, in Maimonides' version, could be other than they 
are, are the size and shape of the world and its parts, the location of the world, the natural place of 
the elements. Maimonides also enumerates features not mentioned by Juwaynl or Shahrastani. 

'"Guide, I, 73(10). 

2m Guide, I, 74(5), in conjunction with I, 73(10). 

201 Above, p. 191. Maimonides and RazI were contemporaries and were separated, geographically, 
by half a world. The argument very likely was circulating in the schools, and Maimonides may have 
learnt it there. 

202 That is to say, something might in theory be possibly existent and eternal, although the advocate 
of creation would not believe that anything in reality is such. 

203 Sec Avicenna's discussion of the "possibly existent by virtue of itself, necessarily existent by 
virtue of another," below, p. 292. 



( 



194 Particularization Arguments 

assumes from the start that everything possibly existent is noneternal and has 
been created. The argument thereby begs the question and is invalid. 204 

4. Ghazali and Maimonides 

Ghazali and Maimonides develop particularization arguments that are purged 
of Kalam elements. Ghazali's formulation is put forward in his Tahafut al-Falas- 
ifa, a work confronting Aristotelian philosophy — as recast by Avicenna — in its 
own terms. Maimonides, for his part, had no sympathy with the Kalam view of 
the universe and subscribed to the Aristotelian view, at least as far as the sublunar 
realm is concerned. Both philosophers accordingly sever the particularization 
mode of thought from its Kalam seedbed and transplant it to an Aristotelian 
framework by seeking out features of the world which are, even from an Aris- 
totelian standpoint, unamenable to natural or rational explanation. 

Ghazali allows himself to be led to his formulation by a dialectical exchange 
of a kind he enjoyed. The mise en scene is the refutation of an old proof for 
eternity, the proof maintaining that no given moment in an empty infinite time 
could, to the exclusion of other moments, have lent itself to the creation of the 
world and recommended itself to the creator as the proper moment for bringing 
the world into existence. 205 As Ghazali sums up the contention to be refuted, 
nothing "could differentiate one specific time [for the world's coming into exis- 
tence] from earlier and later times, seeing that it would not have been impossible" 
for the world to come into existence "earlier or later." 206 The core of Ghazali's 
refutation goes back to Augustine and Philoponus; the creator, he responds, would 
have been able to choose a moment for creation by the exercise of will. 207 But a 
proponent of eternity might, Ghazali recognizes, surrejoin that the exercise of 
will is never a matter of pure arbitrariness, that will invariably opts for what it 
sees as the preferable alternative; and when the alternatives are alike in every 
respect, neither the faculty of will nor any other "attribute" of an agent can 
"differentiate" one alternative from another. The creator's will would thus not 
be able to fix upon a given moment to the exclusion of wholly similar alternative 
moments, and the world must after all be eternal. To counter the surrejoinder, 
Ghazali attempts to convince the advocates of eternity that they cannot help 
acknowledging instances in the universe of the "particularization of one thing 
over what is similar." 208 

He begins by remarking that God's will is seen to decide upon "whiteness to 
the exclusion of blackness" and assign the former to a subject although "the 
subject is as receptive of blackness as of whiteness"; and that God's will is seen 

*»Guide, I, 74(6). 205 Above, p. 53. 

^Tahafut al-Falasifa (Beirut, 1927), I, §28; English translation in Averroes' Tahafut al-Tahaful, 
trans S. van den Bergh (London, 1954), p. 18. 



207 
208 



Above, p. 69. 

Tahafut al-Falasifa, §34; English translation, p. 21. 



Particularization Arguments 



195 



to decide that a body should be in "motion" or "rest" although the body is no 
less receptive of the alternative. By the same token, God's will would have been 
able to decide upon one moment for creation lo the exclusion of the identical 
alternative moments. 209 At this stage Ghazali is apparently working from the old 
Kalam presupposition that when an individual object in the world is black or 
white, in motion or rest, its color and state are not determined by natural forces 
within the universe. The illustrations he adduces, "whiteness and blackness, 
motion and rest," are moreover redolent of the classic Kalam arguments for the 
existence of accidents as actual entities. 210 No adversary, Ghazali knew, was 
likely to be convinced by a rebuttal that presupposes an occasionalistic picture 
of the universe and that provocatively employs illustrations of so dubious a 
pedigree. He therefore quickly passes on to more satisfactory evidence of the 
"particularization" of a given characteristic to the exclusion of equivalent 
alternatives. 

The world, he observes, has a "particular shape" although it might equally 
have alternative shapes. Hence the shape it does have must be selected out arbi- 
trarily by the cause of the world from among the alternatives it might have; and 
by the same token, a moment might arbitrarily be chosen by the cause of the 
world for creation. 2 " Yet an obstinate adversary, Ghazali realizes, might still not 
be convinced. The adversary might retort here that if the world had a different 
shape, if it were "smaller or larger" than it is, or if it had, for example, a different 
"number of spheres and . . . stars," the "universal order" of the world would be 
impaired. 2 ' 2 In other words, although the overall structure of the world, unlike 
the characteristics of individual objects, is not traceable to immanent natural 
forces, it is perhaps explicable on rational grounds as the optimum structure. And 
whereas the cause of the universe can, the adversary will hold, make a rational 
choice and opt for a superior alternative in preference to an inferior, no agent 
can choose arbitrarily between wholly identical alternatives. The upshot of the 
antagonist's retort to Ghazali's rebuttal would again be that a creator could not, 
through the exercise of will, have selected a moment for creation from among 
infinite identical and indifferent alternatives. 

Ghazali at last is driven to the stage he planned to reach from the outset. He 
discovers two features of the celestial region where no natural explanation is 
feasible and where, in addition, there can be no basis for rational, as distinct 
from arbitrary, choice. The two features arc the directions in which the celestial 
spheres move, and the location of a pair of points in the outermost sphere which 
function as poles and around which the spheres rotate. Analogous items had been 
adduced by Juwayni and perhaps by the Ikhwan al-Safa' as evidence of arbitrary 

209 Ibid., §28. I am altering the order of some of Ghazali's statements. 
210 Above, pp. 180, 182. 

2,1 Ibid., §35. 1 have followed the reading given by Tahafut al-Tahafut, I, p. 41. 
2,2 lbid., §36. 



196 Particularization Arguments 

particularization. 213 But Ghazali, unlike his predecessors, endeavors to show how 
these characteristics of the world do evince arbitrariness. 
As regards the second of the two features, he writes that since all parts of the 
( celestial sphere are, on the astronomical theory of the day, alike in nature, nothing 

could recommend any one pair of opposite points to the exclusion of another pair 
as the location of the poles. Therefore, an act of "particularization of one thing 
over what is similar" must be acknowledged, that is, a decision nowise dictated 
by the merits of the alternatives, but completely arbitrary. As regards the first of 
the two features, Ghazali concedes that the rotation of the outermost celestial 
sphere in a direction contrary to that in which the remaining spheres rotate is not 
yet evidence of arbitrariness; the rotation of the spheres in contrary directions 
may be an optimum arrangement for having the influences of the heavens inter- 
mesh and act on the earth in the most productive manner. Instead, his contention 
is that the selfsame result could be achieved in a universe that would be the 
( inversion of ours, in a universe where the highest sphere rotated to the east rather 

than to the west, as it does in the Ptolemaic and medieval astronomical systems, 
while the remaining spheres rotated west rather than cast. Since the movement 
of the spheres in the directions they do move has nothing to recommend it over 
the reverse arrangement, it surely is evidence of arbitrariness and of an act of 
particularization between identical, indifferent alternatives. Ghazali's conclusion 
is that even the proponents of eternity must admit instances in the universe of 
particularization between indifferent alternatives. They cannot, accordingly, object 
to creation on the grounds that no moment might have lent itself to creation to 
the exclusion of another; for the particularizing agent who arbitrarily selected the 
location of the poles and the direction of the movements of the spheres could 
have selected a moment for creation. 214 

Ghazali, in sum, tacitly concedes that the individual characteristics of individ- 
ual objects within the sublunar region do not furnish convincing evidence of 
arbitrary particularization. Such characteristics can be ascribed to the workings 
of immanent natural forces. After restricting his attention to structural features 
of the world, Ghazali explicitly concedes that features rendering ours the best of 
all possible worlds do not constitute convincing evidence of arbitrary particular- 
ization. They may reflect rational, not arbitrary, choice by the cause of the world. 
In the end Ghazali allows his attention to be narrowed to two features of the 
celestial realm which he views as wholly indifferent and as containing nothing 
whatsoever to recommend them over the alternatives. These, he insists, can only 
be explained on the thesis that the cause of the universe selected them by an 
exercise of sheer will. Ghazali's aim throughout is the modest one of refuting a 
familiar proof for eternity by showing that the creator would be able to select a 
moment for creation. He might, however, have framed an independent argument, 



2l3 Above, p. 187. 

2 "Tahafut al-Falasifa, I, §§37-41. 



Particularization Arguments 



197 



akin to the argument of Juwaynfs, 215 wherein evidence of arbitrary choice proves 
that the world is dependent on a voluntary agent, and the dependence of the world 
on a voluntary agent is found to entail creation. Such an argument would go 
beyond JuwaynT in completely divorcing itself from Kalam physics. Maimonides 
rethinks Ghazali's reasoning and docs formulate a particularization argument for 
creation wholly within an Aristotelian, as distinct from a Kalam, framework. 

Maimonides' argument is offered for the sole purpose of proving creation and 
not the existence of God. His rationale in not utilizing the argument for the latter 
purpose apparently is that the particularization procedure takes its departure from 
nothing positive and plays instead upon the inexplicability of various features of 
the physical universe. There could be no way of absolutely precluding an eventual 
evolution of human science to a level where the features in question might be 
explained. As a consequence, an argument from the concept of particularization 
lacked, for Maimonides, the probative weight of the proofs he does offer for the 
existence of God, they being "demonstrations" (burhdn). A demonstration takes 
its departure from true and certain premises and proceeds to an indisputable 
conclusion. Since Maimonides was confident that a demonstration of the existence 
of God was available to him, he would not have knowledge of the existence of 
God rest on anything less. As regards creation, however, where no apodictic 
demonstration was available, he advances the strongest argumentation at his 
disposal. 216 

Maimonides studiously opposes his use of the particularization concept to the 
use made by the Kalam. He expressly rejects the Kalam doctrine of the "atom 
and the continual generation of accidents"; and he refuses to draw any inference 
from the characteristics of things in the sublunar realm, since they can be con- 
strued as "particularized through the powers of the sphere . . . just as Aristotle 
taught us." Characteristics of objects within the world are, that is to say, "parti- 
cularized" in the naturalistic sense of having been determined by physical forces, 
and not in the more significant sense of having been selected through an act of 
will. In contradistinction to Kalam writers, Maimonides undertakes to establish 
the presence of "particularization where it should be established," and to do that 
through "philosophical premises derived from the nature of what exists." 217 The 
domain where he believes particularization, in the significant sense, can be estab- 
lished is the celestial region; and two aspects of the celestial region attract his 
attention. The first is the circumstance that the movements of the spheres do not 
fall into a regular pattern. The second is the location of the stars and planets in 
their several spheres. 

Whereas Ghazali concerned himself exclusively with the nine main celestial 
spheres, Maimonides looks at all the spheres, primary as well as subordinate, 

215 Above, p. 190. 
2i6 Guide, I, 71; II, 16; 19. 
217 Ibid., II, 19. 



198 / Particularization Arguments 

which were repgnized by ancient and medieval astronomy. As many as fifty- 
odd spheres were sometimes hypothesized in order to accomplish the task ancient 
and medic vay astronomy set for itself, the task of reducing to motion in circles 
around the earth what modern astronomy represents as an elliptical motion of 
the planets siround the sun. Each planet, it was theorized, is embedded in a sphere 
that rotates at the surface of another, and it, at the surface of still another. Each 
"Oftfte primary and subordinate spheres in the system was assumed to rotate with 
a constant velocity of its own, but some had to be assumed to rotate with a greater 
and some with a lesser velocity than others; and some had to be assumed to move 
in one direction, while others move in another. What, Maimonidcs asks, might 
account for the diverse velocities and directions in which the spheres rotate? He 
knows that the diversity cannot be explained through a diversity of material 
substance. The spheres, Aristotelian science taught, have a common material 
substance, which expresses itself — or, viewed from the opposite angle, is symp- 
tomized — by something common to them all, not by anything diverse; the com- 
mon material substance of the spheres expresses itself in, and is symptomized 
by, the general circularity of motion common to all the spheres. The diversity in 
velocity and direction must then be due to a factor apart from the substance of 
the spheres. If a regular pattern were detectable in the sequence of spherical 
motions, a uniform natural explanation would be feasible whereby each celestial 
sphere determines the motion of the succeeding sphere. The outermost sphere 
could, for instance, be understood to communicate a proportion of its motion to 
the next sphere, the latter, to communicate an analogous proportion of its motion 
to the next, and so forth. But no pattern was detectable which would permit such 
a hypothesis; for in the supposed sequence of the spheres, westward and eastward 
movements, rapid and slow movements, appeared to succeed each other hap- 
hazardly. Inasmuch as the directions and velocities of the motions of the spheres 
fail to disclose any regular pattern, they can, Maimonidcs insists, be subsumed 
under no necessary natural law. The sole tenable thesis is that a voluntary agent 
"particularized each sphere with whatever direction and velocity of motion he 
wished." 218 

The second feature of the structure of the heavens engaging Maimonides' 
attention is the location of the stars and planets in the spheres. The spheres, on 
the enlightened medieval consensus, rotate constantly, whereas the stars and 
planets imbedded in the spheres undergo no motion of their own. Maimonides 
deduces herefrom that the substance of the spheres must be radically unlike the 
substance of the stars and planets. In addition, the supposed fact that the spheres 
are transparent whereas the stars are luminous also proves to Maimonides that 
the substance of the spheres is radically different from the substance of the stars 

2 ™Guide, II, 19. Elsewhere in the chapter, Maimonidcs brings in the forms of the spheres. He 
writes: A particularizing agent must be posited "who particularized the substrata |of the spheres] and 
prepared them to receive the diverse forms [of which the several motions of the spheres are an 
expression]." 



Particularization Arguments 



199 



and planets. No law of nature and no principle of regularity can, Maimonides 
argues, explain how stars of one substance come to be imbedded in spheres of a 
completely different substance. And no law of nature or principle of regularity 
can account for something still "more extraordinary," namely, the distribution of 
stars in the sphere assumed to contain all the fixed, or true stars, as distinct from 
the wandering stars or planets. In some areas of the sphere of the fixed stars, 
stars of varying magnitudes were seen to be clustered in constellations; in sonic 
areas, stars were seen to be scattered at random; and still other areas were devoid 
of stars. Since the presence and distribution of the stars in their several spheres — 
like the motions of the spheres — exhibit no regular pattern and can be subsumed 
under no necessary law of nature, they too, Maimonidcs maintains, are explicable 
only as an act of choice by a "particularizing agent" who "intended" that the 
stars should be located where they are. 219 

The evidence of particularization adduced by Maimonidcs is unmistakably 
related to the evidence that Ghazali eventually settled upon. As did Ghazali, 
Maimonidcs cites the movements of the spheres and the location of objects — 
planets and stars in his case, poles in Ghazali's — within the spheres. Maimonides, 
though, has introduced a change in that he deems not the arbitrariness, but the 
irregularity of the phenomena to be crucial. The reason for the change can be 
gathered from a totally unrelated context, a theological discussion where Mai- 
monidcs incidentally deals with the selection by the deity of one from among 
several indifferent possible alternatives. In situations of the sort, he writes, it is 
meaningless to ask "why one possibility and not another came to pass; for an 
identical question would occur if the other possibility had come to pass instead 
of the one that did." 220 A choice in any event being called for, no weight can be 
attached to the circumstance that the deity happens to have selected one alternative 
over another. The consideration that when alternatives arc wholly similar and a 
choice has to be made, it is senseless to ask why a given alternative was preferred 
to another, had much earlier been brought to bear in the debate over eternity and 
creation. Philoponus had put forward that consideration in connection with the 
very proof for eternity which Ghazali was rebutting. Faced with the question 
how a given moment could have recommended itself to the creator as the moment 
for creation in preference to other similar moments, Philoponus explained that 
once it is known that the world was created the question why a certain moment 
was chosen becomes meaningless; for an identical question might be posed if 
another moment had been chosen. 221 The same consideration presumably pre- 
vented Maimonides from inferring a "particularizing" agent who acts through 
"intention" cither from the circumstance that the movement of the heavens in 



2,9 ibid. 

22n Guide. HI, 26. Maimonides is dealing with the question why Scripture prescribes the sacrifice 
of a certain number of animals rather than a larger or smaller number. 

221 Above, p. 69. Ghazali also makes the point, Tahafut al-Falasifa, I, §34; English translation, 
p. 21. 



200 Particularization Arguments 

their present directions has nothing to prefer it over their moving in the reverse 
directions, or from the circumstance that the location of the poles in their present 
positions has nothing to prefer it over their location at any other pair of opposite 
points. Seeing that one of the two indifferent sets of directions and one from 
among numberless indifferent pairs of points had to be selected anyway for the 
world to exist, it is, in Maimonides' view, meaningless to ask why a given 
alternative happens to have been selected over another. 

The noteworthy aspect of the movements of the spheres and the locations of 
the stars within them is, for Maimonides, not their indifference but their irreg- 
ularity. If the movements and dispositions of the heavenly bodies formed a nec- 
essary scheme, the scientific observer would expect a certain ratio between the 
velocities of the inner spheres and their distances from the rapid outermost sphere, 
he would expect some affinity between the substance of the planets and stars and 
the substance of the spheres in which they are imbedded, and he would further 
expect that the fixed stars be evenly spaced throughout the heavens. Inasmuch as 
these phenomena lack regularity, they cannot be subsumed under any natural law 
and can be accounted for only on the thesis of a particularizing agent who selected 
them. The selection, it should be mentioned, is not thought by Maimonides to 
be ultimately "purposeless" or, despite the appearance, "haphazard." Believing 
as he does that the universe is the work of an intelligent cause, he is sure that a 
hidden rationality underlies the irregularity exhibited by the celestial realm. 222 

Having come this far, Maimonides adds that "particularization" — or "inten- 
tion," or "choice," or "will" in the proper sense 223 — implies the ability of the 
particularizing agent to control the outcome of his choice. A man cannot, for 
example, be described as "particularizing" himself in, or as "intending" to pos- 
sess, the shape of a being with "two eyes" and "two hands," since he cannot help 
possessing that shape. "Particularization" and "intention" are, consequently, con- 
ceivable solely in connection with "something nonexistent that has the possibility 
of both existing and not existing as intended and particularized"; they thus go 
hand in hand with a product's coming into existence after not existing. Maimon- 
ides has not forgotten that the Arabic Aristotelians spoke of an eternal emanation 
of the universe through an eternal exercise of will on the part of the deity, 224 but 
he dismisses statements of the sort as verbal legerdemain. If the world exists^, 
eternally, it exists necessarily, necessity and eternity being mutually implicative 
by virtue of an Aristotelian principle that Maimonides does not explicitly reforl 
to but undoubtedly had in mind. 225 And to suppose that "existence througn 
necessary emanation" can be combined with a thing's existing by "intention arte! 



221 Guide. II, 19. Cf. above, p. 176. 

223 At the end of Guide, II, 20, Maimonides writes that an agent may be described as willing his 
effect merely in the sense that he has pleasure in it. 
224 Cf. above, pp. 60, 75-76. 

225 Aristotle, De Generalione el Corruption II, 1 1, 337b, 35 ff.; Metaphysics VI, 2, 1026b, 27. 



Particularization Arguments 



201 



will" would be very "c^rjse ... to combining two contraries," thereby violating 
the law of contradiction. The world either exists eternally and necessarily, or it 
exists by particularization, intention, and will, and has come into existence after 
not existing. Given the evidence of "particularization," "choice," "will," and 
"intention" in the celestial region, Maimonides concludes that the celestial region 
and the rest of the world, which is dependent on the celestial region, cannot be 
eternal but must have come into existence. 226 To be precise, the form of the 
physical universe is what must have come into existence. Maimonides' argument 
says nothing about the matter of the universe and does not pretend to be a proof 
of creation ex nihilo. 

Resume 

Kalam writers employed the particularization concept to support their inference 
of a creator from creation. They argued that a particularizing agent must be 
posited who selected out a time for the world to come into existence, who selects 
a time for individual objects to come into existence, who selected existence for 
the world in preference to nonexistence, or who selects out the characteristics of 
each object coming into existence. The contention that a particularizing agent 
must select a full complement of characteristics for each object coming into 
existence borders on a proof for a cause of the world wherein the premise of 
creation is set aside; for it could be argued that whether the world is created or 
not, a particularizing agent must be posited who selects out a complement of 
characteristics for the world and its parts. 

Besides serving to support the inference of a creator from creation, the par- 
ticularization concept found a vr.riety of applications, most notably in arguments 
for the existence of accidents as real entities. A particularizing factor, it was 
reasoned here, must be present in an atom or body to tie each characteristic to 
the given atom or body to the exclusion of the alternative atoms and bodies that 
might have the characteristic, and to tic the characteristic to the atom or body at 
a given time to the exclusion of the alternative times when the characteristic 
might be there. In the course of resolving a difficulty in the theory of accidents, 
Juwaynl indicated why the accident as a particularizing factor is not sufficient 
and a particularizing agent must be assumed as well; the agent must be posited 
to explain how the accident, the particularizing factor, comes to be assigned to 
one atom to the exclusion of others. Juwaynl's explanation again borders on a 
proof for the existence of a cause of the world which sets aside the premise of 
creation. The world, it could be argued, stands in need of an external agent who 
assigns accidents to atoms and bodies. 

Juwaynl, in one of his later works, did formulate a proof of the existence of 
God from the concept of particularization which dispenses with the premise of 



'Guide, II, 20. 



202 



Particularization Arguments 



creation. Since the parts of the world and the world as a whole might, he reasoned, 
have alternative characteristics, an agent must exist who selects out the particular 
characteristics that the world and its parts do have from among all those they 
might have; and a particularizing agent capable of selecting out characteristics 
for the world would have to possess the properties of a deity. Similar arguments 
for the existence of God were offered by other Kalam writers, Juwaynl also 
formulated a new argument for creation. The agent who arbitrarily selects out a 
set of characteristics for the world must, he argued, act through will; but anything 
produced through the exercise of will exists only after not having existed; the 
world, therefore, being the product of an exercise of will, must have cofTle' into 
existence after not having existed. Similar arguments for creation likewise appear 
in other Kalam writers. The new arguments for the existence of God and creation 
contrast with the standard Kalam procedure for proving the same doctrines. The 
standard procedure inferred a creator from creation. The new procedure for prov- 
ing the existence of God derives a cause of the world directly from the presence 
in the world of characteristics that might be different from what they are. And 
the new procedure for proving creation derives creation from the dependence of 
the world on a voluntary, particularizing agent; it concludes with, rather than 
commences with, creation. 

Ghazali had a further application for the particularization concept, employing 
it in his refutation of one of the familiar proofs for eternity. He endeavored to 
show that even the advocate of eternity must acknowledge the ability of the cause 
of the world to make arbitrary decisions and, hence, arbitrarily to have chosen a 
particular moment for the world to come into existence. Maimonidcs, finally, 
developed a particularization argument for creation in a wholly Aristotelian 
framework. He sought out features of the celestial region which cannot be sub- 
sumed under any necessary natural law and which are therefore explicable solely 
as the result of particularization and intention. Whatever is produced by parti- 
cularization and intention cannot, he agreed with the Kalam thinkers, be eternal, 
whence it follows that the celestial region together with the rest of the world must 
have come into existence. 

Such were the principal arguments turning on the particularization concept. A 
progression is to be discerned as well in the phenomena adduced as evidence of 
particularization. The Kalam writers envisaged a world in which the character- 
istics of individual objects are not determined by natural forces and they could 
discover arbitrary selection in the individual characteristics of individual objects. 
Juwaynl's combined proof of the existence of God and creation cited, no doubt 
deliberately, both the characteristics of individual objects within the world and 
structural features of the world as evidence of arbitrary selection. Ghazali tacitly 
conceded that the individual characteristics of individual objects do not constitute 
convincing evidence of particularization, since they can be explained through the 
workings of natural forces; and he explicitly conceded that structural features of 
the universe rendering ours the best of all possible worlds are also not convincing 



Particularization Arguments 



203 



evidence of arbitrary selection. I lc discovered the evidence of sheer arbitrariness 
which he required in features of the celestial region having nothing at all to 
recommend them over the alternatives. Maimonidcs, finally, understood that 
where a choice must be made in any event, no especial significance can be 
attached to one option's happening to be preferred over the equivalent alterna- 
tives. What was crucial for Maimonidcs was not arbitrariness but irregularity in 
the structure of the world. Irregularity would be incompatible with necessary 
natural law and would consequently imply an exercise of choice. 

5. Additional arguments for creation in Maimonides and Gersonides 

Maimonidcs advances two more arguments for creation, without however 
enumerating them as two distinct and coordinate arguments; what I call the first 
of the two appears as a brief appendix to the second. The additional arguments 
arc bracketed by Maimonides with the argument of his examined in the previous 
section for the reason that the new arguments again find the structure of the 
physical universe to be explicable not by any necessary law, but solely as the 
work of a voluntary agent. The first of the additional arguments has, moreover, 
an inner resemblance to the argument already examined. That argument looked 
at the movements of the several celestial spheres and contended that the move- 
ments of the spheres cannot be subsumed under a necessary, natural law; the new 
argument looks at the spheres' forms and contends that these cannot be subsumed 
under any necessary law. Since the movements of the spheres are related to their 
f orms — movement being an expression of form — the new argument can be regarded 
as an extension of the other. 

The first of the additional arguments is concerned with the matter as well as 
the form of the celestial spheres, and with the interaction between celestial matter 
and celestial form. A central Aristotelian thesis had held that objects in the 
celestial region have a different material substratum from sublunar objects. Evi- 
dence therefor was drawn from the supposed fact that the natural motion of 
sublunar objects is rectilinear whereas the natural motion of the celestial spheres 
is circular. 227 According to Maimonides' reading of Aristotle and according to 
what he understood to be the best scientific description of the universe, the world 
contains no less than three separate types of matter: Sublunar objects have their 
common matter, as is indicated by the fact that their natural motion is rectilinear; 
the celestial spheres have another matter, as indicated by the fact that their natural 
motion is circular; and the stars have still a third matter, as indicated by the fact 
that, unlike both the celestial spheres and sublunar objects, they perform no 
motion of their own. 228 Celestial objects, Maimonides further understood, do 
resemble sublunar objects in a certain respect, insofar as they do contain the 



De Caelo I, 2. 
'Guide, II, 19. 



204 



Particularization Arguments 



distinction of matter and form. 229 The celestial spheres arc thus understood by 
him to have a matter common to them and different from the matter of the other 
objects in the universe; and each individual sphere consists in the union of the 
common matter of the spheres with the unique form of the particular sphere. The 
stars similarly have a matter common to them and different from the matter of 
both spheres and sublunar objects; and each star consists in the union of the 
common matter with the unique form of the particular star. 

John Philoponus had long since called attention to an anomaly in Aristotle's 
thinking regarding the material substratum of the heavens. In the Aristotelian 
scheme, the material substratum of objects in the sublunar world constantly 
exchanges one form for another, whereas the substance of the heavens is unchang- 
ing. But if, Philoponus remarked, matter as conceived in the Aristotelian system, 
is "adapted to receive all [possible] forms," the matter of the celestial region 
should be so adapted; it should be "adapted to receive every one of the forms of 
the celestial [bodies]." Just as matter in the sublunar world can, and does, exchange 
one sublunar form for another, each portion of matter in the heavens should be 
capable of exchanging its form for any of the forms suitable to celestial matter. 230 
Philoponus' comment was made as part of a proof for creation which proceeded 
in an entirely different fashion from the proof Maimonides is to offer. His con- 
tention was that since neither the matter of the sublunar world nor the matter of 
the celestial region can by nature retain a form permanently, "nothing [composed) 
of matter and form" can be "indestructible," and the physical universe cannot 
have existed from eternity. 231 

A nearer and more probable source of inspiration for Maimonides' proof is a 
passage in Ghazali's compendium of Avicenna's philosophy. The composition of 
the celestial region is once more the subject of discussion. Ghazali makes the 
unusual statement that the spheres do not after all have a common material sub- 
stratum, that each sphere has a matter peculiar to itself. 232 Very likely at the back 
of his mind was the thought that if each sphere has a matter peculiar to itself, an 
answer is at hand to the question why the substratum of celestial objects is not 
seen to exchange one form for another: Perhaps no portion of celestial matter 
ever assumes a new form because only a single form is suitable for each unique 
portion of matter. 233 The intimation would be that should celestial objects indeed 



This was Avicenna's reading of Aristotle; Averroes, by contrast, rejected the notion that the 
spheres contain the distinction of matter and form. See citations in H. Wolfson, Crescas' Critique 
of Aristotle (Cambridge, Mass., 1929), pp. 594-598. 

230 As cited by Simplicius, Commentary on the Physics, ed. H. Diels, Commentaria in Aristotelem 
Graeca, Vol. X (Berlin, 1895), pp. 1329-1330. 

"'Above, p. 92. 

112 Maqasid al-Faldsifa (Cairo, n.d.), p. 247. I could not find the theory Ghazali is presenting in 
Avicenna's Arabic works, but did find it in his Danesh Nameh; see he Livre de Science, trans. M. 
Achena and H. Masse (Paris, 1955), p. 193. 

233 See an analogous notion in Juwayni, above, p. 184. 



Particularization Arguments 



205 



possess a single common matter, as Aristotle had maintained, the substratum of 
each celestial object might exchange its form for any of the other possible forms 
known, by observation of the heavens, to be appropriate for celestial matter. 

The problematic character of the substratum of the heavens and, it may be 
ventured, the intimation in Ghazali's statement 234 suggest a new proof of creation 
to Maimonides. The essence of matter, Maimonides argues, is to receive, in 
succession, every form appropriate to it. If the spheres are all constituted of a 
common matter, the material substratum of each given sphere is as adapted to 
receive the form of any other sphere as it is to possess the form of its own sphere. 
By the same token, if the stars are constituted of a common matter, the material 
substratum of each given star is as adapted to receive the form of any other star 
as it is to possess the form of its own star. Should events take their natural course, 
the matter of the spheres and of the stars would behave as the matter of the 
sublunar region does, repeatedly shedding one form and adopting another. The 
matter of each sphere would shed its form and successively adopt all the possible 
forms of spheres, with the changes manifesting themselves through repeated 
changes in the direction and velocity of the sphere's motion. The matter of each 
star would likewise shed its form and successively adopt all the possible stellar 
forms, the changes here manifesting themselves by changes in the quality of the 
light radiated by the star. But events plainly do not take their natural course. 
Neither the matter of the spheres nor the matter of the stars behaves as matter 
should behave, successively receiving all the forms it is adapted for. 

The situation perceived by Maimonides is, then, as follows: The spheres have 
the common characteristic of moving circularly, and that should betoken a com- 
mon matter. The stars have the common characteristic of not undergoing any 
motion of their own, and that should also betoken a common matter. Yet neither 
the common matter of the spheres nor the common matter of the stars behaves 
as a common matter should, by nature, behave. In neither case does the common 
matter receive all the forms it is adapted to receive. Since the situation does not 
lend itself to a natural explanation, Maimonides submits his nonnatural expla- 
nation. A conscious voluntary agent must permanently assign a specific form to 
the material substratum of each individual sphere, a form expressing itself in a 
distinctive circular motion; and it must assign a specific form to the material 
substratum of each individual star, a form expressing itself in a distinctive radiated 
light. 235 Maimonides could as well have used the language of particularization 
and written that a particularizing agent must have assigned a particular form to 
the material substratum of each sphere and star. 



234 In Guide, II, 22, Maimonides mentions, as something preposterous, the possibility that "some- 
one should assert that the matter of each sphere is different from the matter of the others." 

235 Maimonides apparently means that God produces each sphere or star by permanently assign- 
ing — through an act of will — a unique form to a portion of the matter common to all the spheres or 
all the stars. 



206 



Particularization Arguments 



Having again come this far, Maimonidcs concludes as before: Eternity implies 
necessity and is incompatible with the action of a voluntary agent. The "intend- 
ing" agent who "particularized" each sphere and star in its fixed form cannot 
therefore have acted from eternity; and the body of the heavens together with the 
sublunar region, which is dependent on the heavens, must have been created. 236 
Nothing has been said regarding the creation of the matter of the universe, and 
the argument is not a proof of creation ex nihilo. 

Maimonides' remaining argument for creation is set forth against the back- 
ground of a theory of emanation which had been fathered upon Aristotle. Aris- 
totle, as most medieval Arabic philosophers read him, bore a Ncoplatonic guise. 
That is, he was understood to have recognized a first cause not merely of the 
motion, but also of the existence, of the universe; and he was understood to have 
maintained that the first cause emanates the universe continually and eternally. 
Aristotelian philosophy in its Neoplatonic guise had, accordingly, to face a per- 
ennial Neoplatonic problem. It had to explain how a highly complex universe 
can flow out of an absolutely simple first cause, considering that, as ttie formula 
went, "from one, only one can proceed." 237 

Alfarabi and Avicenna propounded a solution to the problem, a solution that 
probably was rooted in late Greek Neoplatonism. Alfarabi explained: The first 
cause, which consists in pure thought, 238 eternally emanates its effect by the mere 
act of thinking. Since it has a single object of thought, namely itself, what it 
emanates is a single being; and the latter, flowing from an incorporeal being that 
consists in pure thought, is likewise incorporeal and consists in pure thought. 
There is nevertheless a respect in which the incorporeal being that is emanated 
differs from the incorporeal first cause that emanates it. Whereas the first cause 
has one object of thought, the emanated being has two, namely itself and its 
cause. And inasmuch as the second being has two thoughts, it, through the mere 
act of thinking, emanates two things. It eternally emanates a celestial sphere and 
an additional incorporeal being consisting, again, in pure thought. The additional 
incorporeal being similarly has two objects of thought, itself and the first cause; 
and it too emanates two things, a second celestial sphere and yet another incor- 
poreal being consisting in pure thought. The eternal process is repeated over and 
over, each incorporeal being emanating two effects, until the emanation of the 
active intellect, which is the last of the beings consisting in pure thought, and 
the sublunar corporeal region, which is the last stage of corporeal existence. 239 



Guide, II, 22. There is a certain irony in the argument: It is just the unchanging character of 
the stars and spheres that proves, according to Maimonides, their nonetcrnity. 

237 Avicenna, Shifd': Ilahiydl, p. 405; Ghazali, Maqasid, p. 218; idem, Tahdful al-Faldsifa, III, 
§29, English translation, I, p. 104. Cf. Plotinus, Enneads, V, 1, 6; V, 2, 1; V, 3, 15. In Avicenna 
and Ghazali, the term for "proceed" is yujad or yasdur. 

238 Cf. Aristotle, Metaphysics XII, 9. Plotinus maintained that the first cause, the One, is above 
thought, but exactly what he meant thereby is open to interpretation. 

239 Alfarabi, K. Ard' Ahl al-Madina al-Fadila, ed. F. Dicterici (Leiden, 1895), p. 19; German 



Particularization Arguments 



207 



Such was the scheme of emanation promulgated in the name of true Aristoteli- 
anism by Alfarabi. Avicenna added a nuance and was able to distinguish not two, 
but three thoughts in each of the incorporeal emanated beings. One of the three 
thoughts, in Aviccnna's version, gives rise to the next incorporeal being in the 
series; the second thought gives rise to the soul of a celestial sphere; and the third 
gives rise to the body of the sphere. 240 

The theory was subjected to a harsh critique by Ghazali and, following him, 
by Judah Hallcvi, 241 the aim of both being to expose the depths of absurdity into 
which philosophy can fall. Ghazali registers a number of considerations, two of 
which are germane here. If the first cause is completely free of composition and 
has no more than a single thought, its thought of itself, the being emanated from 
it should, Ghazali objected, be equally free of composition and have no more 
than a single thought, the thought of itself. The first emanated being could in its 
turn give rise to no more than a single additional being, a being that for its part 
would still have just a single thought. The emanation theory of Alfarabi and 
Avicenna might therefore account at most for a series of unitary incorporeal 
entities, but not for the actual physical universe in all its complexity. 242 Further- 
more, Ghazali objected,, even granting that the first emanated being thinks as 
many as three thoughts, each of which gives rise to a new entity, the three thoughts 
could not explain the complexity of the actual physical universe. One of the 
thoughts in the first emanated being is supposedly the source of the outermost 
sphere. But a single thought could not be sufficient for the task, since the sphere 
is complex, containing "form and matter" as well as secondary aspects. 243 And 
the thought in the second incorporeal being which supposedly produces the sec- 
ond celestial sphere would certainly not suffice, seeing that the second sphere 
contains not merely aspects of its own but a thousand-odd stars to boot. 244 

Ghazali's objections, the aim of which was negative, are transformed by Mai- 
monides into an argument for creation. The principle at the heart of the issue is 
ascribed by Maimonides to "Aristotle and everyone who has philosophized," and 
the pregnant formulation Maimonides gives the principle reads: "From a simple 



translation, with pagination of the Arabic indicated: Der Musterslaal, trans. F. Dicterici (Leiden, 
1900). Cf. H. Davidson, "The Active Intellect in the Cuzari and Hallevi's Theory of Causality," 
Revue des eludes juives, CXXXI (1972), 356. 

""Shifd': Ilahiydl, p. 406. In Avicenna 's version, each emanated incorporeal being, or intelli- 
gence, is possibly existent by virtue of itself, necessarily existent by virtue of its cause, and accord- 
ingly has the following objects of thought: itself insofar as it is a possible being; itself insofar as it 
is a necessary being; the first cause. 

241 Hallevi, Kuzari, IV, 25; Davidson, "The Active Intellect in the Cuzari,' p. 358. 

2 " 2 Tahdful al-Falasifa, III, §§37-40, 49; English translation, pp. 109-110, 139. 

243 The secondary aspects are the size of the sphere, and the location of the poles around which 
the sphere rotates; cf. above, pp. 195-196. 

2ii Tahdfut al-Faldsifa, HI, §§54-59; 65-66; English translation, pp. 142-145, 149. 



208 Particularization Arguments 

(' 

thing only one simple thing can necessarily proceed (lazima)." 245 Maimonides' 
argument is to be that on the hypothesis of necessary emanation no satisfactory 

I account of the actual universe is feasible. For the process of necessary emanation 

from a simple incorporeal cause could never give rise to a composite effect; and 
even granting that it could do so inasmuch as the emanated being has two distinct 

( thoughts, duality of thought is far removed from the manifold composition exhib- 

ited by the universe. 

1 Maimonides writes: Though the hierarchy of emanated beings flowing out of 

V the simple first cause should descend "through thousands of stages," no stage 

would possess a greater degree of composition than the preceding one, and the 
I "last [emanated being] . . . would be simple," exactly like the first cause. The 

( process of necessary emanation from an absolutely simple being could, conse- 

quently, never produce a composite effect. But further, conceding for the sake 
of argument that the second incorporeal being docs contain a duality of thought, 
( the full complexity of the universe is unexplained. Maimonides makes the point 

in several ways. He contends: (a) In necessary emanation "a correspondence 
' always obtains between the cause and the effect" so that "a form cannot proceed 

( necessarily from matter, nor a matter from form." Hence an incorporeal being, 

conceding the duality of its thought, could produce only things of the same kind, 
I in other words, incorporeal beings and not a corporeal celestial sphere; and 

{ corporeality in the universe remains unaccounted for. Moreover, granting once 

more for the sake of argument that an incorporeal being could emanate something 
f corporeal, the emanation scheme of Arabic Aristotelianism is still inadequate. 

For (b) a single one of the two thoughts of the supernal incorporeal beings could 
not emanate a full-blown celestial sphere. A celestial sphere consists of no less 
( than four distinct factors, the matter and form of the sphere and the matter and 

I form of the star imbedded in the sphere. On the assumption of necessary ema- 

nation, the thought that is the source of the celestial sphere would itself have to 
f contain four distinct aspects to serve as the source of the four factors in the 

( sphere. 246 Finally (c) the complexity of the sphere of the fixed stars would remain 

totally unaccounted for. That sphere contains stars of various types, and each 
C type is comprised of two factors, its matter and its form. The single thought from 

^ which the sphere is assumed to emanate could not give rise to all those factors. 

The actual complexity and corporeality of the universe are thus completely at 



( odds with the principle affirming that "from a simple thing only one simple thing 

^ can necessarily proceed." The principle, as comprehended and formulated by 

Maimonides, appertains, however, exclusively to necessary emanation. An incor- 
( poreal cause acting, by contrast, not through necessity but voluntarily could 

^ "'Guide, II, 22. 

, 246 As it happens, one of Hallevi's criticisms of the emanation theory was that each succeeding 

incorporeal being would have more objects of thought than the preceding incorporeal being, inasmuch 
( as it has more entities above it; and as a consequence, each should have a larger number of emanated 



product than the previous one. See Kuzari, IV, 25. 

r 
f 



Particularization Arguments 



209 



produce composite as well as simple, and corporeal as well as incorporeal, effects. 
The corporeality and the composition of the universe, which are inexplicable on 
the assumption that the first cause produced the universe through a process of 
necessary emanation, can therefore be satisfactorily explained on the contrary 
assumption that the first cause acted voluntarily. And since a voluntary agent 
does not act eternally, since it acts after not having acted, the world must, Mai- 
monides concludes, be created. 247 Elsewhere Maimonides stresses that the sole 
conceivable kind of causation attributable to incorporeal beings is the process of 
emanation. 248 Here, then, he is advocating a theory of voluntary, nonetcrnal 
emanation in which God initiated the process at a given moment through the 
exercise of his will; God switched on the emanation process, as it were, and 
brought the entire universe into existence. 

The argument, as will be noted, is designed to establish that both the matter 
and the form of the world were created; it is an argument not merely for creation, 
but for creation ex nihilo. 

Gcrsonidcs framed an argument for creation which, although thought out dif- 
ferently from Maimonides' arguments borrows from them. Repeatedly in the 
history of medieval philosophy, one encounters the contention that anything hav- 
ing a cause for its existence cannot exist from eternity; and that the world, since 
it does have a cause of its existence, cannot be eternal, but must have been 
created. 249 The argument Gersonides offers is a refinement of the contention. He 
docs not maintain unqualifiedly that anything with a cause of its existence cannot 
be eternal. What he espouses is the narrower thesis that a subsistent entity with 
a cause of its existence cannot be eternal. 

Like some of the writers already met, notably Juwayni and Maimonides, Ger- 
sonides works from a prior proof of the existence of God, his being a teleological 
argument— to be examined more fully in the next chapter — which establishes 
that the physical universe has "come into existence" and is the "product of an 
agent." 250 Proving that the world was produced by an agent is something less, 
Gersonides makes clear, than proving that the world was created; for, he under- 
stands, there are two distinct ways in which an agent might bring its effect into 
existence. The existence of the effect may, in the first place, be the fruit of "an 
absolute coming into existence" ; that is to say, the effect may be endowed with 
existence "solely at the moment of its coming into existence," as when a "house" 
is produced by a "builder." 251 But existence may, in the second place, be "ema- 
nated continually" from the cause, as when "motion" is produced continuously — 



2,1 Guide, II, 22. 

"'Guide, II, 12. 

249 Above, p. 190; below, p. 387. 

250 Below, p. 231. 

15l Milhamot ha-Shem (Leipzig, 1866), VI, i, 6, p. 309; 7, p. 312. 



210 



Part icularizai ion Arguments 



in the pre-Newtonian universe — by a cause of motion. Should something be 
brought into existence in the first way, should it all at once acquire an existence 
that it did not previously have, it obviously cannot have existed from eternity. 
Should, however, something be emanated in the second way, it may well — 
although it not necessarily will — have existed from eternity, since the cause of 
its continual emanation may always have emanated it. An illustration of this 
possibility is the motion of the spheres, which is "emanated" from the "thinking 
of the [incorporeal] movers of the celestial spheres"; inasmuch as the motion of 
the spheres is emanated continually, the process might be eternal, the movers of 
the spheres continually and eternally bringing about the spheres' motion. 252 

Having laid down the distinction between things that are brought into existence 
once and for all and things that are emanated continually and perhaps eternally, 
Gersonides reduces the question of creation to another question. His teleological 
proof for the existence of God had established that the celestial region and, 
through it, the rest of the world were brought into existence by a cause. He now 
asks whether the cause bestowing existence on the celestial region did so once 
for all or whether it does so through a continual and perhaps eternal process. 
Ruling out the latter hypothesis will prove that the world was created. 

Consider, writes Gersonides, the hypothesis that the celestial region is ema- 
nated from the deity continually. The hypothesis is certainly not viable if its 
purport is that the deity sustains the heavens in existence by, at each moment, 
converting already existent celestial bodies into new, exactly identical celestial 
bodies. On such a construction the celestial region would exist unchangingly and 
unchanged, "no act at all" would ever occur, and to speak of a cause of the 
existence of the heavens would be nonsensical. The hypothesis that the celestial 
spheres are emanated by their cause continually can have meaning only on the 
supposition that "immediately upon the spheres' being emanated from God, they 
are destroyed [and return] to nothingness, whereupon they are immediately re- 
emanated . . . from God out of nothing." But the supposition that the heavens 
are constantly being destroyed and rcemanatcd is on its face ludicrous; and analy- 
sis, further, shows that it embodies a logical absurdity. The heavens would, on 
that supposition, not just come into being and forthwith be destroyed. If gaps are 
not incessantly to interrupt their existence, the heavens would have to come into 
existence, be destroyed, and come into existence again at the very same moment. 
The heavens would thus have to be both existent and nonexistent at every moment 
in time, a situation precluded by the law of contradiction. 253 

The supposition that the spheres are constantly emanated, destroyed, and re- 
emanated has, moreover, bizarre corollaries. Time would not be continuous, but 
would consist of the contiguous discrete moments at which the heavens exist. 
The heavenly region would have potential, and not actual, existence, because 



Ibid., 6, p. 308; 7, p. 314. 
Ibid., 7, p. 313. 



Particularization Arguments 



211 



each discrete episode in the existence of the spheres would last no more than a 
moment, and "as is proven in [Averroes' commentary on] the Physics, anything 
existing for merely a moment exists potentially, not actually." 254 And the heavens 
would be unable to undergo continuous motion, or for that matter any motion 
whatsoever. For a new set of heavens would be reemanated at each moment, and 
no set of heavens would exist long enough to move. 255 

It is to be noted that Gersonides' reasoning, although spelled out in conjunction 
with a distinctive picture of the celestial region, is not wedded to any physical 
scheme. Gersonides is attempting to refute the entire notion of an entity's being 
produced through continuous emanation. An entity produced through continuous 
emanation would, he reasons, either have to exist forever unchanged; and to 
speak of its being caused would then be nonsensical. Or else the entity would 
have to be constantly destroyed and recreated, which is ludicrous and ultimately 
illogical. As a "general rule," he summarizes, "things that endure cannot possi- 
bly" be emanated continually and, if caused, must be brought into existence once 
for all; "only accidents that do not endure, such as motion" can be emanated 
continually, and they alone might be emanated from eternity. 256 As regards the 
heavenly bodies, since they are subsistent entities and do have a cause of their 
existence, their cause cannot be of the sort that emanates its effect continually, 
but must be of the sort that brings its effect into existence at a single moment. 
Neither the celestial region nor the sublunar region, which is dependent on it, 
can therefore have existed from eternity. 257 

Gersonides buttresses his conclusion through an auxiliary argument, obviously 
inspired by Maimonides, that details the "things found in the heavens" which 
are incompatible with the "nature of the heavens." From the "physical sciences" 
he knows that the substratum of all the heavenly bodies is of a "single . . . nature 
. . . containing no diversity whatsoever." 258 If the heavenly bodies are all con- 
stituted of exactly the same homogeneous substratum, they should be completely 
alike; and in fact no more than a single celestial body should exist. Yet not merely 
docs more than one heavenly body exist, the heavenly bodies differ from each 
other in a number of respects. They are, Gersonides finds, endowed with different 
forms, as can be inferred from the differing influxes descending from them into 
the sublunar realm. The spheres are of different sizes, even though size is, by 



254 See Averroes, Middle Commentary on Physics, (Oxford, Bodleian Library, Hebrew MS., 
Ncubauer 1380 = Hunt. 79), VIII, v, 3, pp. 94a-b, referring to Aristotle, Physics VIII, 8, 263a, 
23 ff. 

2 "Milhamot ha-Shem, VI, i, 7, pp. 312-313. 

2,6 Ibid., p. 314. Gersonides' thinking presumably is: Since an accident such as motion is not 
subsistent, new motion neither is produced out of the previous motion, nor does it require the previous 
motion to be destroyed. 

257 Ibid. 

""DeCaelol, 2-4. 



212 



Particularization Arguments 



nature, a function of the "mixture" of matter (mezeg) in an object, 259 and the 
celestial bodies share the same material substratum. The spheres occupy different 
places in space, one encompassing the other, even though objects of the same 
substance should have an identical natural place. 260 The spheres differ in the 
number of stars they contain, the sphere of the fixed stars containing a thousand- 
odd stars, other spheres 261 containing one each, and still others containing none. 
The heavenly bodies differ insofar as some, the spheres, are transparent, whereas 
others, the stars and planets, are luminous and nontransparent. The stars and 
planets, for their part, differ from each other in their colors and in the diverse 
influences they exercise upon the earth. These differences, Gersonides submits, 
although wholly inexplicable from the standpoint of the nature of the celestial 
substratum can be accounted for on the thesis that the celestial region was brought 
into existence through "will and choice." The upshot of his proof is that the 
celestial region as well as the sublunar realm, which is dependent on it, must 
have been brought into existence after not having existed, by a transcendent agent 
who exercised will and choice. 262 



259 As Gersonides explains, the mixture of matter in an ant determines a certain size and that in a 
camel determines another size. The minor variations from the norm in the size of an ant or a camel 
are due to their being "compounded from contraries," and even such minor variations should not be 
exhibited by the spheres, since their matter does not contain contraries. 

260 In other words, they should not stand one above the other, but all should occupy exactly the 
same place, at the same distance from the center of the earth. 

26l Those containing the wandering stars, or planets. 

262 Milhamot ha-Shem, VI, i, 8, pp. 316-320. 



VII 

Arguments from Design 



1. Cosmological, teleological, and ontological proofs 
of the existence of God 

The preceding chapters have dealt primarily with arguments for eternity and 
creation, but several, relatively simple proofs of the existence of God have also 
been met. The latter are all associated with the Kalam tradition, though not in 
every instance thought out in a specifically Kalam framework. The most popular 
of the proofs of the existence of God met thus far is the one resting on a prior 
proof of creation: Given the creation of the world, the conclusion is drawn that 
an agent must have been responsible for the world's coming into existence. The 
weight in that procedure falls upon the preliminary proof for creation. Once the 
creation of the world is established, the existence of a creator is inferred, either 
as something self-evident 1 or by analogy with objects observed to come into 
existence within the world, 2 or through the subtler thought, turning on the concept 
of particularization, that an agent would be required to select out the particular 
moment for the world to come into existence as well as other particular charac- 
teristics for the world when coming into existence. 3 Besides the proof from 
creation we have met a proof of the existence of God from the concept of par- 
ticularization in which the premise of creation is dispensed with; and proofs of 
the existence of God from composition in the universe. The proof from the 
concept of particularization dispensing with the premise of creation takes its 
departure from features of the physical universe which might conceivably be 
other than they arc; and it reasons that since the universe could as well have had 
alternative characteristics, an agent must be posited who selected out the char- 
acteristics the universe does have. 4 Proofs of the existence of God from com- 
position in the universe reason either in general that the composition of the 
universe indicates an intelligent and powerful cause, 5 more narrowly that the 

'Above, pp. 154-155. 
2 Abovc, pp. 156-157. 
3 Above, pp. 161, 178. 
"Above, p. 187. 
'Above, pp. 148-149. 



213 



214 



Arguments from Design 



composition of contrary qualities in the universe indicates such a cause, 6 or else 
that the design evidenced in the composition of the universe must be due to such 
a cause. 7 Arguments establishing a cause of the universe are not in themselves 
complete proofs of the existence of God, although the Kalam thinkers unfortu- 
nately do not state what, precisely, a complete proof must comprise. After arriving 
at a cause of the universe, adherents of the Kalam proceed to argue that the 
ultimate cause of the universe is eternal, 8 one 9 and incorporeal, 10 possessed of 
power, knowledge, life, and will" — but they do not tell us precisely which of 
these attributes are being established as part of their proofs of the existence of 
God and which are ancillary attributes. 

The distinction between ontological, cosmological, and teleological 12 argu- 
ments for the existence of God is always illuminating. Of the proofs met thus 
far, all but one are cosmological, insofar as they fit Kant's paradigm exactly, 13 
but in the broader sense that they reason from the existence of the world or from 
the existence of something in the world, to a cause of the world. One proof, that 
from the design evidenced by the composition in the universe, is. teleological. 
None is ontological. None, that is, arrives at the existence of God purely through 
an analysis of concepts. 

The chapters to follow will have as their subject medieval Islamic and Jewish 
proofs of the existence of God which are associated with the Aristotelian tradition. 
These are: the Aristotelian proof from motion; an offshoot of the proof from 
motion which has been called the proof from 'logical symmetry'; Avicenna's 
proof of thcexistence of a being necessarily existent by virtue of itself; and a 
family of arguments from the impossibility of an infinite regress of causes. All 
of them take their departure from the existence of something, be it motion in the 
world, an individual object in the world, or, less frequently, the world as a whole. 
And they reason therefrom to the existence of a first cause. They are, accordingly, 
like the majority of proofs associated with the Kalam tradition, cosmological. 

One of the proofs associated with the Aristotelian tradition might, if read 
carelessly, be misinterpreted as an ontological rather than a cosmological argu- 
ment. Central to Avicenna's proof of the existence of a being necessarily existent 
by virtue of itself is the analysis of the critical concept, the concept necessarily 
existent by virtue of itself. And a superficial reading might lead to the misappre- 
hension that the existence of a being corresponding to the concept is derived by 

6 Above, p. 151. 
7 Above, p. 152. 

"Ghazali, Tahafut al-Falasifa, ed. M. Bouyges (Beirut, 1927), HI, §§3, 16; English translation 
in Averroes' Tahafut al-Tahafut, trans. S. van den Bergh (London, 1954), pp. 89, 96; above, p. 165. 
'Above, pp. 166-171. 

'"Above, pp. 171-172. "Above, p. 173. 

I2 1 employ the expressions argument from design and teleological argument as synonyms, although 
by etymology an argument showing order without purpose and functionality should not, perhaps, be 
called teleological. Kant's name for the argument from design was the "physico-theological proof." 

l3 See Kant, Critique of Pure Reason, A604 - 606/B632-634. 



Arguments from Design 



215 



Avicenna solely from an analysis of the concept. 14 The error might, moreover, 
be abclled by the presence in European philosophy of ontological arguments for 
the existence of God which do consist exclusively in the analysis of a similar 
concept, that of necessary being. 15 Avicenna's proof, it turns out, docs not arrive 
at the existence of a being necessarily existent by virtue of itself solely through 
analysing a concept, and his proof is unambiguously cosmological. 16 

Were a desperate need felt to unearth a specimen of ontological argumentation 
in medieval Arabic philosophy, a more promising area to search would be not 
the writings of Avicenna himself, but the minor philosophic works that his writ- 
ings inspired. It might, for example, be imagined that a passage in the 'Uyun al- 
Masa il offers an ontological argument. The 'Uyun al-Masa'il, a text sometimes 
wrongly ascribed to Alfarabi, is a philosophic fiorilegium, a catena of excerpts 
and paraphrases from philosophic writings, especially from Avicenna. One para- 
graph in the text reads: "To assume that the necessarily existent docs not exist is 
impossible. Its existence has no cause, and cannot be (?) from another." 17 Should 
the statement — which is based on Avicenna's definition of the necessarily existent 
by virtue of itself 18 — be taken as a self contained proof of the existence of God, 
the intent could only be that the mere inspection of the concept necessarily 
existent renders the denial of the existence of a being corresponding to the concept 
absurd. The passage would then constitute an ontological argument. But that 
construction is militated against by the failure of the author, or compiler, of the 
'Uyun to intimate that his statement is in fact intended to stand as a self-contained 
proof of the existence of God. Immediately prior to the passage, a brief cosmo- 
logical proof had been given, and now the author apparently is saying: Once a 
necessarily existent being is known to exist through the preceding argument, such 
a being cannot fail to exist at all times; for it has no cause and is dependent on 
nothing else. Placing an ontological construction on the passage in question is 
also militated against by the absence of any suggestion that the author or compiler 
had a notion of what ontological argumentation might be. The 'Uyun al-Masa'il 
may be safely judged free of the ontological argument of the existence of God, 
and, more comprehensively, medieval Islamic and Jewish philosophy may as a 
whole be judged free of ontological argumentation. 

14 Scc Jak. Gultmann, Die Religionsphilosophie des Abraham ibn Daud (Goctlingcn, 1879), p. 
121; H. Wolfson, "Notes on Proofs of the Existence of God in Jewish Philosophy," reprinted in his 
Studies in the History of Philosophy and Religion, Vol. I (Cambridge, Mass., 1973) p. 569; P. 
Morcwcdgc, "Ibn Sina and Malcolm and the Ontological Argument," The Monist, LIV (1970), 238 
and 242. 

"See below, pp. 392-393. 

"'Sec below, pp. 298, 303-304, 403-404. 

"In Alfarabi's philosophische Abhandlungen, ed. F. Dietcrici (Leiden, 1890), p. 57; German 
translation, with pagination of the Arabic indicated: Alfarabi's philosophische Abhandlungen ausdem 
Arabischen uebersetzt, trans. F. Dietcrici (Leiden 1892). Other examples arc R. Zaynun, in Rasa il 
al-Fiirabi (Hyderabad, 1931) pp. 3-4, and al-Da'awa al-Qalbiya, pp. 2-3, in the same volume. 

l8 Cf. below, p. 291. 



216 



Arguments from Design 



Most proofs of the existence of God associated with both the Kalam and 
Aristotelian traditions are, to repeat, cosmological. The sole telcological argu- 
ment encountered until now is the proof from the design evidenced by the com- 
position of the universe. Additional arguments from design can, however, be 
gleaned from the writings of the Islamic and Jewish philosophers. The arguments 
have differing purposes; they seek to prove the existence of God or to establish 
certain attributes in God. Teleological arguments are more common among writ- 
ers standing in the Kalam tradition, but they appear as well among those who 
stand in the Aristotelian tradition and those who would eschew labels. In the 
medieval, as in the ancient, period, teleological arguments are not closely rea- 
soned. An effort was virtually never made to determine precisely what constitutes 
evidence of design, the supposition being that anyone who sees design will intu- 
itively recognize it. Furthermore, arguments that are ostensibly proofs of the 
existence of God generally fall short of establishing the existence of a being 
possessed of all the specifications of the deity and therefore remain something 
less than complete proofs of the existence of God. A critique of teleological 
argumentation has been preserved from the ancient period, but it is rudimentary, 19 
nowise approaching the critique of Hume, for example. 20 I was able to discover 
no more than a soupcon of a critique in medieval Islamic and Jewish literature. 21 
It may be of the nature of teleological arguments that they are the least sophis- 
ticated proofs of the existence of God and yet, when all has been said, remain 
the most plausible. 

2. Teleological arguments 

The teleological arguments preserved from ancient philosophy pursue one of 
two lines, discovering evidence of design either on a minor, or a cosmic scale. 
The evidence of design which is discovered is either the functionality of nature 
or an aesthetic quality in nature, its orderliness and beauty; and there is a partial 
correlation between the two lines of argumentation and the two kinds of evidence 
which are adduced. Arguments on the minor scale pass in review individual 
details from various realms of nature and usually focus, specifically, on the 
functionality of each detail. Arguments conducted on the cosmic scale may sim- 
ilarly adduce the overall functionality disclosed by nature as evidence of design, 
but to a greater extent they concern themselves with the aesthetic quality, the 
orderliness and beauty, of the cosmic panorama. 

Xenophon relates that the first of the two lines of argumentation constituted a 
part of Socrates' positive teaching. Socrates, the report goes, would point out 
the functionality of divers aspects of man's physiological and psychological 

"See Cicero, De Natura Deorum, III, ix-x, 24 - 25; xi, 28 
20 Cf. Hume, Dialogues concerning Natural Religion, II. 

2l Judah Hallevi, Kuzari, I, 19-20. Hallevi himself employs teleological arguments; see below, 
pp. 228-229. 



Arguments from Design 



217 



endowments and also phenomena in the world surrounding man which contribute 
to man's well-being. He postulated that what "clearly is for a purpose" must be 
the result of "forethought" (-yvcoixt)), not "chance." And he concluded that man 
must have been made by a "wise demiurge"; and again that the "gods take care 
to furnish men with the things they need," that nature discloses divine "provi- 
dence" and "love of man." 22 The same procedure was employed, in an expanded 
form, by the spokesman for Stoicism in Cicero's De Natura Deorum. More data 
arc offered there; the inquiry is extended to phenomena of botany, zoology mete- 
orology, and geology, which have nothing to do with man; and the wealth of 
evidence discloses that functionality and hence design are ubiquitous in nature. 
The conclusion of the spokesman for Stoicism is that "the world is governed by 
the providence of the gods" — that "everything in the world is governed, for the 
welfare and preservation of all, by divine intelligence and deliberation," and not 
"by chance." 23 The style of reasoning employed by Socrates and by the Stoic 
was likewise used by Galen, who uncovers the intelligent operation of "nature" 
or the "demiurge" in numerous details of human and animal physiology. 24 

The second line of teleological argumentation in the ancient period, the one 
that searched for design on a cosmic scale, directed its attention in some instances 
exclusively to the heavens; and what was cited was the utility of the heavens, 
their immense beauty, and the regularity of celestial motion. The motif goes back 
to Plato's Laws 25 — and indeed to the Book of Isaiah. 26 It was expressed elo- 
quently by the younger Aristotle in his dialogues. 27 And it was enunciated several 
times by the spokesman for Stoicism in Cicero's De Natura Deorum. The "uni- 
formity" of celestial motion and the "utility, beauty, and order" of the heavenly 
bodies prove to the spokesman for Stoicism that a "mind . . . governs" the 
universe. 28 In other instances the aim of the argument was even more ambitious 
and a cosmic plan was discovered not merely in the heavens, but in the universe 
as whole. An argument of the sort was employed, once more, by the spokesman 



"Xenophon, Memorabilia, I, iv; IV, iii. Scholars have debated whether the sections in question 
arc correctly attributed to Xenophon and whether the account that is given of Socrates' teachings is 
accurate. 

"Cicero, De Natura Deorum, II, xxix, 73; xlvii, 120 — lxvi, 167; and more briefly in Tusculan 
Disputations, I, xxviii, 68. 

24 Galen, De Usu Parlium, X, 9; XVII. The theme had also appeared in Aristotle's De Partibus 
Animalium, but there design in nature is clearly regarded as immanent and nonconscious. See De 
Purtihus Animalium I, 1, 641b, 12; III, 1, 661b, 24; III. 2, 663b, 23-24; IV, x, 687b. 

"Plato, Laws, 966-967; cf. 897-899. 

26 Isaiah 40:26. Cf. also Quran 71:15-16. 

"Cicero, De Natura Deorum, II, xxxvii. 95 (designated as Aristotelian fragment #12); Sextus, 
Adversus Physicos, I, 20-22 (fragment #10). These passages are assigned by scholars to the lost 
dialogue De Philosophia. 

2 "De Natura Deorum, II, v, 15; sec notes of J. Mayor and A. Pease in their editions, ad locum. 
The theme also appears in De Natura Deorum, II, ii, 4; xxxviii, 97; xliv, 115; and in Sextus, Adversus 
Physicos, 1. 26. 



218 



Arguments from Design 



for Stoicism in Cicero, 29 and Philo. 30 As expressed in the language of Stoic 
philosophy, the "sympathetic, harmonious, all-pcrvading affinity of things . . . 
forces" one to recognize "a single divine, all-pcrvading spirit." 11 Whether it is 
in the heavens alone or in the universe as a whole that an all-embracing design 
was sought, the ancient philosophers had a repertoire of analogies at their dis- 
posal. They compare the heavens or the entire universe to a house, 32 a wrestling 
school, 33 a deliberative assembly, 34 a city, 35 a ship, 36 a book, 37 or a piece of 
machinery such as a clock. 38 No impartial observer encountering any of these 
objects and perceiving how well ordered it is, how well its parts mesh, and how 
well they function, would — so the contention went — imagine that the object 
before him had come about spontaneously. No one would imagine that it could 
have arisen by pure chance, that it is not the handiwork of an intelligent designer. 
By the same token, no impartial observer gazing upon the heavens or the universe 
in its entirety can imagine that they came about through pure chance and are not 
the handiwork of an intelligent being who designed them. A certain advantage 
is to be noted in the cosmic line of teleological argumentation, especially from 
the standpoint of medieval thinkers, for whom the unity of God was more fun- 
damental than for the ancients. Teleological argumentation on a cosmic scale 
lends itself to a demonstration of the unity, as well as of the existence, of a cause 
of the universe. If the universe exhibits a single overall design, the conclusion 
is more easily drawn — although it is not necessarily drawn — that a single designer 
is responsible. 

Both lines of teleological argumentation, that from the details of nature and 
that from the overall design exhibited by nature, reappear in medieval Arabic 
philosophy. Usually, though not always, arguments pursuing the first line adduce 
functionality as evidence for design, while arguments pursuing the second, adduce 
the orderliness of nature. That is to say, Arabic arguments from design stress the 
functionality, but sometimes the orderliness or also the beauty of the details of 
nature; or else they stress the orderliness or perhaps the beauty, but sometimes 
the functionality too, of nature as a whole. A number of Ideological considera- 
tions deployed in medieval Arabic literature are akin to those deployed in classical 

29 De Natura Deorum, II, vii, 19; cf. xlv, 115. 
w Legum Allegoria, III, xxxii, 99. 

3l De Natura Deorum, II, vii, 19; cf. notes of Mayor and Pease. 

"Cicero, De Natura Deorum, II, v, 15; Philo, Legum Allegoria, III, xxxii, 98. Sec Pease's note 
to Cicero, ad locum, for further examples. 
"Cicero, ibid. 
34 Ibid. 

"Philo, Legum Allegoria, III, xxxii, 98. 

16 Cicero, De Natura Deorum, II, xxxiv, 87; 89: "the course of a ship"; Philo, Legum Allegoria, 
III, xxxii, 98: "a ship"; Sextus, Adversus Physicos, I, 27: a ship seen travelling across the oceans. 
Further examples in Pease's note to Cicero, ad locum. 

■"Cicero, De Natura Deorum, II, xxxvii, 93. 

"Ibid., xxxiv, 87-88; xxxviii, 97. 



Arguments from Design 



219 



literature, and the resemblance is such that a historical continuity is unmistakable. 
At least one route whereby the transmission was effected can be traced. 

Both lines of teleological argumentation passed into patristic literature, where 
writers use them to prove the existence of God and the pervasiveness of divine 
providence in the universe. 39 As a rule, literary links between the Church Fathers 
and Islamic thought arc not detectable, but in the case of the argument from 
design the stages in the transition can be observed. The links are a Greek work 
by the fifth-century Christian cleric Theodoret, entitled On Providence; and an 
Arabic work, extant in at least three slightly different recensions, which is some- 
times attributed to the well known litterateur Jahiz and sometimes to one Anbari, 
apparently a Christian. 40 The Arabic work appears under varying names, each 
of which contains the term reflection. In the printed edition, where the book is 
attributed to Jahiz, it is called: The Book of Proofs and Reflection regarding 
Creation and Divine Governance (K. al-Dala il wa-l-l'ti'bar 'ala al-Khalq wa- 
l-Tadbir). Thcodoret's treatise On Providence undertakes, as the title promises, 
to set forth the operation of divine providence in the universe. A large amount 
of evidence for design is presented, much of it strikingly similar to evidence 
adduced by the spokesman for Stoicism in Cicero. Since Cicero utilized Stoic 
sources in Greek and Theodoret probably was restricted to Greek materials, 
Cicero and Theodoret very possibly borrowed from the same Stoic work — per- 
haps a lost work of Posidonius' 41 which was still available to the Church Fathers; 



Lactantius, De Opificio Dei; Institutiones. 1, 2; VII, 3; Ps. Clement, Recognition s. VIII, 
chaps. 20, 22, 23, 44; Basil, Hexameron, I, 7; VI. 1 and 10; Sermon on Deuteronomy 15:9, §§7— 
8; Gregory of Nyssa, De Anima el Resurrectione, 111; Augustine, De Civitale Dei, XI, 4; Nemcsius, 
De Natura Hominis, chap. 42; K. Gronau, Poseidonius und die jeudisch-christliche Genesisexegese 
(Leipzig, 1914); M. Pohlcnz, Die Stoa (Gocttingcn, 1955-1959), note to p. 431. 

4 "(1) A'. al-Dala' il wa-t-ftibar 'ala al-Khalq wa-l-Tadbir. attributed to Jahiz (Aleppo, 1928). 
(2) K. al-'Ibar wa-l-I ' tibtir. attributed to Jahiz (British Museum. MS. Or. 3886); described by H. 
(iil)l), " The Argument from Design," Ignace Goldzilur Manorial Volume. Vol. 1 (Budapest, 1948), 
pp. 1 50- 162. (3) K. al-l-'ikr wa-i-l 'tibar, attributed to Jibril b. Null b. Abi Null al-Anbari (Aya Sofia, 
MS. 4836/2. pp. 160-187). A photograph of the Aya Sofia manuscript was made available to me by 
Josef van Ess, and he called to my attention that a certain Abu Nuh al-Anbari, who is presumably 
identical with the grandfather of our Jibril. is known as a Christian; sec P. Krauss, "Zu ibn al- 
Muqaffa'," Rivisla degli Studi Orientali, XIV (1933), 10-11. Militating against the attribution of 
the text to Jibril al-Anbari is a statement in the introduction to the British Museum manuscript to the 
effect that one of four sources used by the author of the text was a poorly written and ill-organized 
book by the same Jibril b. Nuh al-Anbari. See below, n. 42. 

The Aya Sofia manuscript is exactly the same as the printed Aleppo text except for occasional 
changes or omissions of a word or two, and except for occasional missing sentences, which arc 
possibly just marginal glosses that were introduced into the printed text. The body of the British 
Museum manuscript is identical with the body of the printed text. But the British Museum manuscript 
has an introduction and conclusion of its own, each running several pages; and the printed text has 
a long concluding section of eighteen pages which does not appear in the British Museum manuscript. 

41 See J. Mayor's edition of De Natura Deorum, Vol. II (Cambridge, Mass., 1883), pp. xx-xxiii; 
Gronau. Poseidonius und die juedisch-christliche Genesisexegese. 



220 



Arguments from Design 



and at any rate they draw from a common Stoic pool of thought. The K. al- 
Dala'il, for its part, also puts forward evidence of design in the universe, the 
aim here being to establish the existence of a "creator," to establish more precisely 
that the entire universe is the handiwork of a single creator, and to establish that 
faultless providence and wisdom pervade the world. In the introduction to one 
of the recensions of the K. al-Dala'il, Jahiz, or pseudo-Jahiz, names the sources 
he used, and included among them is Theodoret. 42 The body of the Arabic work, 
moreover, evinces parallels with Theodoret's On Providence. The K. al-Dala'il 
is known to have been in circulation in the Middle Ages; for adaptations and 
excerpts have been identified in a number of Arabic writers, both Islamic and 
Jewish. 43 A complete itinerary is thus traceable in which the teleological argu- 
mentation of the ancients can be seen traveling through Theodoret to the K. al- 
Dala'il, whence it diffused into Islamic and Jewish literature. A few illustrations 
may be instructive. 

(a) Cicero reports that Cleanthes and Chrysippus of the old Stoa argued for 
the existence of the "gods" with the aid of an analogy between the design apparent 
in a house and the design apparent in the universe. The evidence they cited is of 
an aesthetic character. Cleanthes contended: "When someone comes into a house" 
and finds everything regulated and arranged systematically, he cannot "suppose 
that these [things] came about without a cause"; with far more reason, the "motions 
. . . and order" of the world, which over an infinite past time have "never . . . 
played false," must convince every observer that nature is "directed by a mind." 44 
Chrysippus compared the "beauty of the heavenly bodies . . . and the great 
magnitude of the ocean and lands" to "a giant and beautiful house"; and he 
inferred that the world must have been constructed by the "immortal gods" as a 
"domicile" for themselves. 45 Philo too was familiar with the analogy and he 
credited it to members of the school "whose philosophy is reputed the best." The 
philosophers in question — presumably the Stoics — maintained: "Should a man 
see a carefully built house ... he will get a notion of the craftsman. . . . Just 
so, anyone entering, as it were, the great house of the world . . . and beholding 
the heavens, . . . planets, stars, . . . which move rhythmically, harmoniously, 
and for the benefit of the whole, . . . beholding as well the [arrangement and 
variety of] earth, . . . water, . . . air, . . . and fruits, . . . will surely conclude 
that these [things]" are the work of "a creator, God." 46 The analogy recurs in the 
patristic writers, 47 and among them is Theodoret, his intent being not so much 



"The passage is published by C. Rieu, Supplement to the Catalogue of the Arabic Manuscripts 
in the British Museum (London, 1894), pp. 466-467; translated by Gibb, "The Argument from 
Design," pp. 153-154. 

"See below, pp. 223-224. "De Nalura Deorum, II, v, 15. 

45 Ibid., vi, 17. 

46 LegumAllegoria, Ill, xxxii, 98-99. Cf. De Specialibus Legibus, I, vi, 33-35. 
47 Lactantius, Institutiones, II, 9; Ps. Clement, Recognitions, VIII, 20; Pease's note to De Nalura 
Deorum, II, v, 15. 



Arguments from Design 



221 



to demonstrate the existence of God as to determine God's nature. "At the sight 
of a very skillfully made house," Theodoret writes, "we forthwith admire the 
craftsman . . . [although] not present, . . . and we attribute the entire beautiful 
form ... to his craftsmanship. Similarly, when we see the heavens and the 
beneficial procession of lights across them, the magnitude and beauty of the 
creatures give us some conception of the creator." 48 And the motif reappears in 
the K. al-Dala'il. Jahiz, or pseudo-Jahiz, states at the beginning of his treatise: 
The "early writers ... did well" when they compared the "world" to "a house 
that has been constructed and made ready, a house containing every appointment," 
while man is "like the householder placed in charge of everything therein." As 
construed by the author of the K. al-Dala'il, the analogy suffices not only to 
establish the existence of an intelligent cause of the universe; on the grounds that 
the world reveals a single unified design, the Arabic text concludes that a single 
creator must have made the world and provide for it. The resemblance of the 
world to a house is a "clear proof that the world is created with governance 
(tadblr), planning (laqdir) ... and order, ... and that the creator of the world 
is one, he being [the agent] who fitted the parts together and arranged them." 49 
(b) Xenophon relates that Socrates cited the functionality of the human speech 
organs as a piece of physiological evidence for divine providence. The "gods," 
Socrates taught, rendered the human tongue "capable of producing articulate 
speech through touching different parts of the mouth at different times," all with 
the purpose of letting men "communicate whatever they wish to each other." 50 
The mechanics of human speech arc subsequently cited as evidence of divine 
providence by Cicero's Stoic. The Stoic speaker sees the wisdom of "God" or of 
"nature" in the tongue, which "renders the sounds of the voice distinct and clear 
by striking the teeth and other parts of the mouth." And here the design inherent 
in the speech organs is brought out through an analogy with a musical instrument. 
"Our school," the speaker remarks, "is wont to compare the tongue to the plec- 
trum [whereby the lyre is plucked]; the teeth to the strings; and the nostrils to the 
horns of the lyre [i.e., to the hollow arms extending from the body of the instru- 
ment]. . . " 5I The comparison of the tongue to the plectrum of a lyre and the 
teeth to the strings became commonplace, 52 and Theodoret was one of the writers 
adopting it. But Theodoret was careful to add a reservation. He recalls that in 
truth art imitates nature, and not nature, art; 53 hence, the art of fashioning musical 
instruments must have imitated the operations of providence and not vice versa. 



Theodoret, De Providenlia, in Palrologia Graeco-Latina, cd. J. -P. Migne, Vol. LXXXIII (Paris, 
1864), col. 608. 

49 K. al-Dala'il, p. 3; cf. p. 63. 
*° Memorabilia, I, 4. 

"De Nalura Deorum, II, lix, 147; 149. On II, vi, 18, Cicero refers to Xenophon's reports 
regarding Socrates' teleological teachings. 

52 See notes of Mayor and Pease to De Natura Deorum, II, lix, 149. 
"See Aristotle, Physics II, 2, 194a, 21-22. 



222 



Arguments from Design 



"In place of the teeth, art has stretched the strings. In place of the lips, art has 
inserted the bronze [sounding board]. And the plectrum serves as a tongue for 
[plucking] the strings." 54 The workings of divine providence in "the human . . . 
voice organs" also engage the attention of the K. al-Dald'il. Its author too knows 
of the analogy with a musical instrument, and once again cites the analogy in the 
name of the "early writers." The comparison now drawn, however, is different 
from Theodoret's, the organs of speech being likened — more appropriately — to 
a wind, rather than to a stringed instrument. The Arabic passage cannot, as a 
consequence, be directly, or solely, dependent on Theodorct. Yet since the thought 
is unchanged, and the remark regarding the relationship between art and nature 
is repeated, the Arabic passage is unmistakably continuing the same tradition. 
According to the K. al-Dala'il: "The [human] throat is like a pipe for producing 
the voice, and the tongue, lips, and teeth [scrvc| to shape the letters and melo- 
dies." The "early writers therefore ... did well" when they "compared the throat 
to the tube of a wind instrument (mizmdr) . . , and drew an analogy between — 
on the one hand — the lips and teeth that shape the voice into letters and melodies, 
and — on the other hand — the fingers that travel over the mouth of the wind 
instrument and shape its piping sounds into tunes." As for the question which is 
imitating which, the analogy with a musical instrument is, writes the K. al- 
Dala'il, legitimate enough "for the sake of demonstration and instruction," that 
is to say, as a pedagogical device for exhibiting the presence of design in the 
world. It must nonetheless be remembered that "in truth the wind instrument 
copies the . . . voice. The instrument is made by art, whereas the voice is natural; 
and art copies nature." 55 

(c) To take an additional example, Socrates is reported to have insisted upon 
the functionality for man not only of the light of day but also of the darkness of 
night. Inasmuch as man "stands in need of rest," Socrates explained, the "gods 
provide [him] with a most beauteous time for resting." 56 The notion is repeated 
by the spokesman for Stoicism in Cicero's De Natura Deoruin. Among the gifts 
of divine providence, the Stoic counts "the alternation of day and night," which 
"affords a time for acting and another time for resting." 57 This motif as well 
became a commonplace. 58 In Theodoret's formulation, divine providence is dis- 
closed in the movement of the sun. "For in rising, the sun produces the day, 
while by setting and, as it were, hiding itself, the sun yields to the night." During 
the day, especially during the long days of summer, men attend to their affairs. 
Yet "night provides men with no less benefit than day," particularly 59 insofar as 



"De Providemia. col. 592. 
"K. al-Dala'il. pp. 50-51. 
56 Xenophon, Memorabilia, IV, 3. 

51 De Natura Deorum, II, liii, 132. The notion also appears in Philo, De Specialibus Legibus, II, 
xx, 100-103. 

5R Sec Pease's note to Cicero, ad locum. 

'"Another function of night, according to Theodorct, is that it prevents men from becoming sated 
with daylight. 



Arguments from Design 



223 



night prompts men to "rest [their] weary bodies awhile," insofar as it compels 
"even exceedingly industrious men to desist from labor." 60 In the same vein, the 
K. al-Dala'il detects the hand of providence in the "rising and setting of the 
sun." The rising of the sun permits men "to busy themselves in their affairs." 
The "usefulness" of the setting of the sun consists, primarily, 61 in furnishing men 
with the opportunity of "repose for the recovery of their bodies," and indeed in 
forcing them, when necessary, to rest; for human greed is so great that "many 
would never rest except for ... the darkness of . . . night." 62 

In addition to the preceding examples, the K. al-Dala'il accumulates hundreds 
of details from plant biology, zoology, human physiology, and human psychology, 
and the treatise sets forth the functionality of each detail; some, though by no 
means all, of the details are akin to those in the classical sources. 63 The "order," 
"permanence," and "purposcfulncss ," evinced by the world as a whole and the 
functionality of the several parts of the world demonstrate to the K. al-Dald'il 
that the world cannot have come about "by accident" and "neglect" (ihmdl), but 
must have been made in accordance with a "plan" by a "creator . . . [who] is 
one." 64 

As already mentioned, material from the K. al-Dald'il can be identified in 
other Arabic texts. A book attributed to Ghazali and entitled The Wisdom in God's 
Creatures (al-Hikma ft Makhluqdl Allah) consists largely of excerpts from the K. 
al-Dala'il. 65 The entire K. al-Dald'il was recast as a dialogue by a Shiite author. 66 



m De Providemia. cols. 565, 568. 

&l Night, according to the Arabic text, also has the function of preventing the world from becoming 
overheated. 

62 K. al-Dala'il. p. 4. The Quran too had cited the alternation of night and day as evidence of 
divine design; see below, p. 227. 

'''The following are a few other examples of continuing motifs: 

The human eye is providentially protected by eyelids, lashes ("a palisade"), and eyebrows ("a 
roof"). Cf. Xenophon, Memorabilia. IV, 3; Aristotle. De Partibus Animalium 11. 15; Cicero, De 
Natura Deorum, II, Ivii, 143; Theodorct, De Providemia, col. 601; Jahiz, K. al-Dala'il. p. 52. 

The digestive and excretory functions show clear evidence of design. Cf. Xenophon, Memor- 
abilia, 1, 4; Cicero, De Natura Deorum, II, Iv, 137-138; Theodorct, De Providemia, col. 612; Jahiz, 
K. al-Dala'il, pp. 46-47, 57. 

Man's erect stature is evidence of divine providence. Cf. Xenophon, Memorabilia, 1, 4; Cicero. 
De Natura Deorum, II, Ivi. 140; notes of Mayor and Pease, ibid.; Jahiz, K. al-Dala'il. p. 47. The 
purpose served by man's erect posture was explained in various ways; see notes of Mayor and Pease. 

The willingness of animals such as the ox, the horse, the dog, and the ass, camel, or elephant 
to submit to man, although they are stronger than he, is evidence of divine concern for man. Cf. 
Xenophon, Memorabilia. IV, 3; Cicero, De Natura Deorum. II, lx, 151. and Ixiii, 158-159; Theo- 
dorct. De Providemia Deorum, cols. 633-637; Jahiz, K. al-Dald'il, pp. 28-29. 
**K. al-Dala'il, pp. 2, 3, 47, 66, 67. 

™ al-Hikma fi Makhluqdl Allah, (Cairo. 1908). SeeD. Baneth, "The Common Telcological Source 
of Bahyc ibn Paqoda and Ghazzali" (in Hebrew with English Summary), Magnes Anniversary Volume 
(Jerusalem, 1938), pp. 23, 28-29. 

" r 'K. al-Tawhid. attributed to Mufaddal b 'Umar and incorporated into al-Majlisi's encyclopedic 
work, Bihar al-Anwdr. See Baneth, "The Common Telcological Source," p. 27, n. 33; J. van Ess, 
Die Gedankenwelt des Harit al-Muhasiln (Bonn, 1961), p. 172. 



224 



Arguments from Design 



Sections were reworked in a treatise written by one Ahmad b. Sulayman, also a 
Shiite. 67 And excerpts are incorporated by the Arabic Jewish writer Bahya into 
his Duties of the Hearts.™ The tcleological theme is fairly frequent in medieval 
Arabic and Hebrew literature, as the examples to follow will show. The theme 
happens, however, to be prominent in the Quran. 69 Consequently, in most of the 
examples to be given, it can only be conjectured how far the teleological mode 
of thought is an outgrowth of the classical tradition, mediated through the K. al- 
Dala'il or another channel, and embraced because of the congeniality of the 
theme; and how far that mode of thought may have sprouted directly from the 
seedbed of the Quran. 

MuhasibI (d. 857) has a teleological argument for the unity of the cause of the 
universe which might serve equally well as an argument for the existence of such 
a cause. Throughout the universe, in inanimate nature, plant life, animal life, 
and human life, MuhasibI discovers that "each part fits together with the others"; 
and a number of details he adduces seem distinctly to echo details in the K. al- 
Dala'il. The interconnections reveal that the universe, from its lowest to its 
highest level, forms "one whole," and the unity of "governance" evidenced by 
the universe leads MuhasibI to infer the unity of the cause of the universe. 70 

A brief teleological argument for the existence of God is given by the Shiite 
author Qasim b. Ibrahim (785-860). The "imprints" of perfect wisdom and the 
"signs" of good governance manifest in the universe prove to Qasim that a wise 
and good deity must be responsible. 71 Maturldl (d. 944) offers a brief argument 
wherein he derives the existence of a "creator" (muhdith) from what he calls the 
"principle" that everything in the universe exhibits "wondrous wisdom." 72 The 
unity of the creator is subsequently inferred from the fact that each process in 
nature — the seasons of the year, the paths of the heavenly bodies, the life cycles 
of plants and animals — observes its own unvarying and uninterrupted course. 73 
Maturldl does not state, but perhaps he meant, that these several processes of 
nature are interdependent and mesh, and hence stand in need of a single, all- 
envisaging architect. In a separate teleological argument for the unity of the cause 
of the universe Maturldl points to the circumstance that diverse species and widely 
scattered individuals belonging to the same species have their needs provided for 
throughout the universe. The interdependence of the workings of providence, he 
concludes, shows a "single . . . provider (mudabbir)" to be responsible. 74 

The Ikhwan al-Safa', the "Brothers of Purity," — who are thought to belong to 
one of the Shiite branches of Islam — offer a teleological argument for the exis- 
tence of God in which the heavens supply the evidence of design. The celestial 

67 Gibb, "The Argument from Design," p. 152, n. 5. 
68 Baneth, "The Common Teleological Source," pp. 28-29. 
69 See Quran 2:164; 3:190; 30:20-25; 45:3-5; 55. 

70 See van Ess, Die Gedankenwelt des Hark [al-Muhasibi, pp. 163-166, 170-171. 
"See W. Madelung, Der Imam al-Qasim ibn Ibrahim (Berlin, 1965), p. 106. 
"Maturldl, K. al-Tawhid (Beirut, 1970), p. 18. 

"Ibid., p. 21. 74 Ibid., pp. 21-22. 



Arguments from Design 



225 



"spheres" (falak) and "stars," the Ikhwan observe, differ in respect to size; and 
the motions of the heavenly bodies differ, moreover, in respect to velocity as 
well as direction, some moving "cast, (some] west, [some] south, (some) north." 
Those data could, as was seen in the previous chapter, furnish the underpinning 
for an argument from the concept of particularization, an argument that detects 
arbitrariness or irregularity in celestial phenomena and proposes to explain the 
arbitrariness or irregularity. 75 But what the Ikhwan al-Safa' detect is design and 
wisdom. The Ikhwan conclude, with no ado, that the ingenious arrangement of 
the celestial spheres cannot have come about by chance but must have "occurred 
through the intention of an intending agent . . . [who is] wise and powerful." 76 

A more circumstantial teleological argument for the existence of God is advanced 
by Ibn Hazm. Ibn Hazm had formulated an elaborate proof of the existence of 
God from creation, and his teleological argument is appended almost as an after- 
thought. 77 He uncovers the requisite evidence of design at two levels, on a cosmic 
scale, in the arrangement of the celestial region, and on a lesser scale, in mis- 
cellaneous details of biology. The celestial phenomena attracting Ibn Hazm's 
attention arc the phenomena that had constituted evidence of arbitrariness or 
irregularity in the aforementioned argument from particularization, but that had 
furnished evidence of design in the passage just quoted from the Ikhwan al-$afa\ 
In working out his argument, Ibn Hazm enters into astronomical technicalities 
to a greater extent than did the Ikhwan. 

The primary theoretical task of ancient and medieval astronomy had been to 
construe the motion of the planets in conformity with the principle that celestial 
motion is circular and the further principle that the earth stands at the center of 
the universe. What modern astronomy sees as an elliptical motion of the planets 
around the sun had, accordingly, to be reduced to motion in circles around the 
earth. Some astronomers explained the motion of the planets through the hypoth- 
esis of eccentric spheres. An eccentric sphere is a sphere whose center is different 
from the earth's, and the hypothesis of eccentric spheres assumes each planet and 
star to be embedded in the surface of such a sphere: The eccentric sphere rotates 
circularly at a constant velocity; the center around which the sphere rotates is 
itself in circular motion; and the interaction of the circular motions was under- 
stood to give rise to the noncircular motion of the planets. 78 Other astronomers 
explained the motion of the planets through the hypothesis of epicycles. Epi- 
cyclical spheres are secondary spheres rotating at fixed points on the surface of 
other spheres — which may themselves rotate at the surface of still others — the 
underlying, major spheres in the system being concentric to, and in constant 
circular motion around, the earth. 79 



'Above, pp. 187, 197-199. 

'Ikhwan al-Safa, Rasail (Beirut, 1957), III, pp. 335-336. 
'Cf. above pp. 120, 122. 

'Cf. J. Dreyer, A History of Astronomy from Thales to Kepler (New York, 1953), pp. 144-146. 
'Ibid., pp. 152-157. 



226 Arguments from Design 

I 

^ Ibn Hazm does not distinguish between the two hypotheses, and in fact they 

are not mutually exclusive. 80 With the characteristics of eccentric spheres in mind, 
/ he marvels at the circumstance that the celestial spheres "have different centers" 

around which they rotate, yet despite their having different centers, they "(it 
1 together tightly" and are able to "maintain their circular motion" and unvarying 

( velocities. 81 With the characteristics of the epicyclical spheres in mind, be mar- 

vels at the "variance" between the "motion of the epicyclical spheres" and the 
' "motion ... of the [main] supporting spheres"; and at the additional circum- 

i stance that the "revolutions of all the Imain] spheres [with the exception of the 

ninth are] from west to east" whereas "the revolution of the ninth [and outermost] 
' sphere," which is responsible for the daily movement of the heavens, is "from 

( east to west." The interaction of the motions of the secondary, epicyclical spheres 

with the motions of the underlying, main spheres, and the interaction of the 
^ eastward motions of all but one of the main spheres with the "conflicting," west- 

I ward motion of the outermost sphere give rise to the intricate courses of the 

planets; and the entire epicyclical arrangement is so ingenious that it must be the 
( fruit of conscious planning. The operations of the eccentric spheres and of the 

I epicycles thus lead "us . . . necessarily" to recognize the hand of a "mover." 82 

When Ibn Hazm turns from the celestial to the terrestrial region, he assembles 
( data from animal and plant biology. Unlike most medieval proponents of teleo- 

| logical argumentation, who, when treating the details of nature, saw only their 

functionality, he underlines the aesthetic side of the details of nature. He admires 
/ the skill by which the limbs of the human body are fitted together, a motif touched 

^ on in the K. al-Dala ;7; 83 the uniform color patterns of sundry "animals, . . . 

i birds, tortoises, reptiles (hashardt), and fish"; the variegated plumages of other 

( species of bird, also touched on in the K. al-Dala' i'/; 84 the fact that palm tree 

■ , fiber has a texture as skillfully woven as fabric from a loom, again a motif present 

; in the K. al-Dala' il. ss Ibn Hazm concludes: It is incontrovertibly "known through 

'I the necessity of intellect" that the celestial and terrestrial regions must have come 

j. about by the "deliberation of a maker" who "exercises choice and invention." 

j And the evidence of design on both the macrocosmic and microcosmic levels is, 

( he avers, sufficient not merely for concluding that the universe has a maker, but 

j, for concluding as well that it has a "single" maker. 86 

A simple teleological argument for the existence of God is put forward by 
I Ghazali in one of his less technical works. Like Ibn Hazm, Ghazali offers the 

'/ 

80 Ibid.. pp. 155-156. 

( 8, Maimonides touches upon problems of this sort in Guide to the Perplexed, II, 24. 

"K. al-FaslJT al-Milal, Vol. I (Cairo, 1964), pp. 18-19; Spanish translation: Abenhdzam dc 
\ Cdrdoba y su Historia Critica de las Ideas Religiosas, trans. M. Asm Palacios, Vol. II (Madrid, 

, 1928), pp. 111-113. Ibn Hazm does not draw the conclusion until after giving his examples from 

} the terrestrial region. 

, S3 K. al-Dala'il, p. 47. 

84 Ibid., pp. 38-39, also in Nemesius, De Natura Hominis, chap. 42. 
( 85 Ibid., p. 23 . 86 Ar. al-FasI ft al-Milal, p. 19. 

( 



Arguments from Design 



227 



teleological argument side by side with the proof from creation, the proof that 
first establishes the creation of the world and then infers the existence of a creator. 
But unlike Ibn Hazm, Ghazali indicates a preference for the Ideological argu- 
ment — which he describes as "inborn" in man and as so evident that "setting up 
a demonstration" is, in reality, superfluous. Ghazali begins by quoting passages 
from the Quran which contain the teleological theme, for example Quran 2: 164: 
"In the creation of the heavens and the earth, in the alternation of night and day, 
in the ship (fulkl) that runs on the sea, ... in the driving forth of winds and 
clouds, . . . there arc surely signs for people who understand." After quoting 
several such passages, Ghazali declares: Nobody "possessing the least intelli- 
gence who reflects upon . . . these vcjrses, who gazes upon the wonder of God's 
creation on earth and in heaven, who gazes upon the marvelous formation of 
animals and plants," can doubt that "the well adapted arrangement" depends on 
a "maker who governs . . . and adapts it." 87 

The teleological argument for the existence of God is also employed by Bahya 
ibn Paquda, a writer who in addition used teleological material borrowed from 
the K. al-Dala' il in contexts apart from the one to be examined now. 88 Once 
again, in Bahya, the teleological argument for the existence of God appears in 
the company of an argument for the existence of God from creation. Here, 
however, the two arguments are not juxtaposed and coordinate. Rather, the argu- 
ment for the existence of God from creation is a more comprehensive train of 
reasoning into which teleological considerations arc incorporated at several stages. 
In the more comprehensive argument, Bahya considers composition in the uni- 
verse. He points out, as was seen in an earlier chapter, 89 that the world is com- 
posite; from composition he concludes that the world was created; and from 
creation he infers a creator. But in the first stage of the argumentation, when 
ostensibly adducing evidence of composition, Bahya makes a comment that can 
stand by itself as a self-contained proof of the existence of God. The thinking is 
that the world as a whole has been designed to accommodate man: and Bahya's 
source is clearly the K. al-Dala' il. He writes: "Looking at the world, we find it 
to be joined and composite . . . like a house that has been constructed and made 
ready in every appointment," while man is "like the freeholder of the house, who 
makes use of everything therein"; for "the intent in each part [of the world] is 
the benefit and well-being of rational beings." 90 At a subsequent stage of the 
argumentation, after having completed the body of the wider proof, after having 
deduced creation from the composition in the universe and inferred the existence 
of a creator from creation, Bahya confronts cavillers who may retort that the 
world has no cause and exists "by chance." His rebuttal turns out to be wholly 



S7 al-Risala al-Qudsiya, published as Al-Ghazali's Tract on Dogmatic Theology, ed. and trans. A. 
Tibawi (London, 1965), Arabic text, p. 16; English translation, pp. 33-34. 

88 See Bancth, "The Common Teleological Source," pp. 28-29; below, p. 235. 
89 Above, pp. 152-153. 

^'al-Hiddya (Hobol ha-Lebabot), ed. A. Yahuda (Leiden, 1912), 1, 6. Cf. above, p. 221. 



228 



Arguments from Design 



independent of the evidence he had adduced for composition in the universe and 
equally independent of the inference of a creator from creation. No one of "sound" 
mind, he submits, would suppose that a device such as the waterwheel whereby 
farmland is irrigated can come about "without the design of a maker." 91 How 
then could the "great wheel" of the heavens, which serves the "well-being of the 
entire earth and everything in it," have come about "without the intent of an 
intending agent and the planning of a wise and powerful agent?" No one, more- 
over, would imagine that a book might have ben produced merely by "spilling 
ink on blank paper"; 92 how then might the world, which is of "an infinitely . . . 
finer craftsmanship," have come about "without the intent of an intending agent, 
the wisdom of a wise agent, and the power of a powerful agent?" 91 The world 
must be the handiwork of an intelligent maker. 

When Bahya takes up the question of the unity of God, he bases that tenet too 
on evidence of design. He writes: The "traces of wisdom in God's creatures arc 
similar and alike," the entire world being governed by "a single order and a single 
motion, which comprehends each and every part." These "traces of the creator's 
wisdom" in "the least and most exalted parts of creation testify that they derive 
from a single wise creator." 94 

A teleological proof diverging from the argumci..s discussed so far is offered 
by Judah Hallevi. Hallevi's proof has its roots in a strain of medieval speculation 
regarding the source of the formal, as distinct from the material, side of natural 
objects. And his is a proof that goes a bit beyond mere intuition, beyond the 
presupposition that no one of sound mind can fail to recognize design in the 
details of the world and in the world as a whole. Form in the Aristotelian sense 
is the inner principle that makes any object what it is, that gives the object its 
essence. The contention of certain medieval philosophers, notably Avicenna, had 
been that the formal side of a natural object is radically different from the inert, 
material side, and that a cause can never give rise to what is totally different from 
itself. The conclusion drawn was that the formal aspect of any natural object 
cannot emerge from within, from the material aspect, but must enter from with- 
out, from a cause consisting, like the effect, in form. The formal aspect of each 
natural object must come to the material substratum from a transcendent giver of 
forms. 95 Against the background of that line of speculation, Hallevi formulates 
a teleological argument. He reasons that the only qualitative traits of physical 

"Sec above, n. 38. 
92 Sce above, n. 37. 

93 al-Hidaya (Hobot ha-Lebabot), I, 6. 

94 Ibid., I, 7 (2, 3, 7). Bahya has no less than seven proofs of the unity of God. The second is 
wholly teleological, and the teleological motif is woven together with other strands in the third and 
seventh proofs. 

95 Sec H. Davidson, "Alfarabi and Avicenna on the Active Intellect," Viator, III (1972), 149- 
150, 156-157; idem, "The Active Intellect in the Cuzari and Hallevi's Theory of Causality," Revue 
lies eludes juives, CXXXI (1972), 368-377. 



Arguments from Design 



229 



objects attributable to the blind forces of nature are the elemental and wholly 
physical qualities, that is to say, heat, cold, wetness, and dryness. By contrast, 
"the instilling of form, the evaluation [of the appropriateness of a given portion 
of matter for a specific form], . . . and whatever involves wisdom [acting] towards 
a goal, can be attributed solely to a wise and powerful agent" who is beyond 
nature. 96 Whenever a natural form appears in a material substratum, the form 
cannot, consequently, emerge from the matter itself but must be the handiwork 
of an external source, operating consciously and towards a goal. Hallevi docs 
not go beyond the affirmation that the world is the product of a "wise and powerful 
agent." His argument therefore falls short of establishing either a single cause or 
a first cause and, like most teleological arguments, cannot pretend to be a com- 
plete proof of the existence of God. 

Another writer who makes use of teleological reasoning is Averrocs, and the 
context in which he does so is illuminating. Averrocs makes clear in several 
works that from a scientific and philosophic standpoint, the only fully adequate 
proof of the existence of God is Aristotle's proof of the existence of a first mover. 97 
In the treatise where he expresses approval of the teleological argument the tone 
and subject matter arc not scientific and philosophic, the subject being, instead, 
the proper method of teaching fundamental truths, particularly the existence and 
unity of God, to nonphilosophers. Methods whereby different theological schools 
and especially the Kalam sought to establish the existence and unity of God are 
passed in review; and the various methods arc rejected by Averrocs on the grounds 
that they are both invalid and at variance with the spirit of the Quran. 98 As 
Averroes interprets the Quran, two arguments for the existence of God are rec- 
ommended there. First, the Quran recommends what is in effect a simplified 
cosmological argument, an argument concluding in a nontechnical fashion that 
some entity must be responsible for the occurrence of events in the world. Sec- 
ondly, it recommends an argument running as follows: "Everything in the world 
is adapted" to the needs of the human species and reveals "providence." "Day 
and night, sun and moon," the earth and everything therein, the organs of the 
human body — all serve the needs of man. The functionality exhibited throughout 
the world cannot conceivably be due to "chance." It must "perforce" be the doing 
of "an agent . . . who intends . . . and wills it"; and the "existence of a creator" 
is thereby established. 99 When Averrocs, in the present treatise, turns to the unity 

96 Kuzari, I, 77. Sec the article in the Revue des eludes juives, cited in the previous note. 

97 Tahafut al-Tahafut, cd. M. Bouyges (Beirut, 1930), pp. 393-394; English translation, with 
pagination of the Arabic indicated; Averroes' Tahafut al-Tahafut; Epitome of Metaphysics, cd. and 
Spanish trans. C. Quiros Rodriguez (Madrid, 1919), IV, §3; German translation: Die Epitome der 
Melaphysik des Averroes, trans. S. van den Bcrgh (Leiden, 1924), pp. 105-106. 

98 A*. al-Kashf, cd. M. Mueller (Munich, 1859), pp. 27-28; German translation, with pagination 
of the Arabic indicated: Philosophic und Theoiogie von Averroes aus dem Arabischen uebersetzt, 
trans. M. Mueller (Munich, 1875). 

"Ibid., pp. 43-44. 



230 



Arguments from Design 



of God, the argument which he finds to be recommended by the Quran is only 
marginally teleological. The Quran, as read by Avcrroes, observes that the uni- 
verse is a single unified effect and from the unity of the effect it infers the unity 
of the cause. 100 As for the incorporeality of God, no argument for it is derived 
by Averrocs from the Quran. In his opinion, the incorporeality of God is not a 
doctrine that should be imparted to common folk and hence not a doctrine with 
which Scripture would deal. 101 

Averroes further states that the two proofs for the existence of God which can 
be extracted from the Quran — the simplified cosmological argument and the 
teleological argument — are exactly the "procedure" philosophers employ when 
proving the existence of God. The philosophic formulation of the proofs and the 
popular formulation differ merely "in degree," the philosophic formulation being 
of greater comprehensiveness and greater "profundity." 102 What Averroes means 
when referring to a philosophic formulation of the scriptural cosmological argu- 
ment is plain. The precise, philosophic formulation of the cosmological argument 
would be nothing other than Aristotle's proof from motion. What he means when 
referring to a philosophic formulation of the scriptural teleological argument is 
evidently to be taken in the same vein. He must be permitting himself a certain 
liberty; and his meaning must be that in a loose sense the proof from motion 
subsumes the teleological argument, and the latter can be thought of as a popular 
version of the former. 103 

Averroes' position, then, is that purposefulness is to be discovered in the 
workings of nature, 104 and the teleological argument contains sufficient grains 
of truth, and is sufficiently plausible, to answer the requirements of nonphilo- 
sophers. The Quran, which addresses an audience of nonphilosophers, accord- 
ingly reasons ideologically. The teleological argument is not, however, a fully 
adequate demonstration of the existence of God, the chief reason presumably 
being that argumentation from the functionality of nature views the universe 
anthropocentrically. The universe is represented as if it existed exclusively for 



""Ibid., pp. 47-48. 
""Ibid., p. 67. 
""Ibid., pp. 46, 48. 

103 Avcrroes can sometimes be seen to change his mind from one work to another; but there is no 
reason to suppose that we have an instance of that in his characterizing the Aristotelian proof from 
motion as the only adequate proof for the existence of God in certain of his works, while apparently 
referring in the present work to a philosophic version of the teleological argument. Averroes' changes 
of mind seem always to concern the proper interpretation of a specific passage in Aristotle and never 
to be conscious departures from Aristotle. 

It is curious that Aristotle offered a teleological argument for the existence of God in one of 
his early popular dialogues (which was lost by the Middle Ages and was unknown to Avcrroes), but 
omitted that form of argument in his later more technical works (see above, n. 27); and now Avcrroes 
characterizes the teleological argument as appropriate for nonphilosophers, whereas in his own tech- 
nical works he recommends only the Aristotelian proof from motion. 

l04 See Aristotle, Physics II, 8; De Partibus Animalium. 



Arguments from Design 



231 



the sake of man, whereas in truth the celestial realm cannot exist "primarily" to 
serve man, since the "superior" never exists "for the sake of the inferior." I(,s The 
sole proof of the existence of God which meets the standards of serious philos- 
ophers would be Aristotle's proof from motion. 

Gersonidcs was highly dependent on Avcrroes and yet he strained constantly 
for originality and independence. As it happens, despite Avcrroes' patronizing 
assessment of the teleological proof for the existence of God, Gersonidcs adopts 
the proof. Gersonidcs' version stands in the tradition of teleological arguments 
that search for a single overall purpose in the heavens. But his version is unique 
in its broad astrological assumptions. 

Gersonidcs begins by laying down "the most distinctive property" of things 
whose existence has a cause, that is to say, the characteristic found in all such 
things and found exclusively in them. "The most distinctive property" of things 
whose existence has a cause, whether they exist by "nature" or "art," is, he 
writes, that "they exist for a certain purpose." Iah In the terminology of the Aris- 
totelian theory of four causes distinguishable in objects — a material cause, formal 
cause, final cause, and efficient cause — Gersonidcs' principle means that any- 
thing disclosing a final cause must also have an efficient cause. 107 Examination 
of the celestial region, Gersonidcs continues, reveals that "everything" in the 
heavens is adapted "in the highest possible degree, to bring terrestrial beings to 
perfection"; and the functionality is so extensive that it cannot conceivably be 
"accidental." The number of stars and planets, their diverse magnitudes, their 
varying distances from earth, their distribution through the heavens, the diversity 
in their radiation, their several movements, arc adjusted and attuned in a way 
that allows the influences descending from the heavens to work the maximum 
benefit for sublunar existence. Since the heavens unmistakably exist for a purpose, 
and since they show evidence of a "final cause," they must belong to the class of 
beings that have an "efficient cause" of their existence. 108 In a separate passage, 

'""•TalHlfitl al-Tahdful. pp. 484-485. 

""'Milhamot lui-Slwm (Leipzig, 1866). VI, i. 6. p. 308. Tor the definition of property, see Aris- 
totle. Topics I, 5, 102a, 18-19; Averrocs, Middle Commentary on Porphyry's Isagoge and on Aris- 
totle's Categoriae, trans. H. Davidson (Cambridge and Berkeley, 1969). pp. 16-17. 

""See Aristotle. Physics II. 7; I-. Zellcr. Die Philosophic der Chechen. Vol. II. Part 2. (4th cd.; 
Leipzig. 1921), p. 328. 

"'"Milhamot ha-Shcm. VI, i. 7. pp. 310-312. Gersonidcs later finds evidence of design and of 
a designer in an aspect of the sublunar world, in the fact that dry land exists although the natural 
place of the element earth is beneath the natural place of the element water. He reasons: The existence 
of dry land obviously serves a purpose; it makes possible the existence of land animals. Given the 
rule that anything existing for a purpose has an efficient cause of its existence, an efficient cause of 
the existence of dry land must be posited. And since the phenomenon is contrary to nature, (hat cause 
must not be a natural, but a transcendent agent. Sec Milhamot ha-Shem. VI, i, 13. p. 350. The motif 
goes back to Slobacus. Geography. XVIII, i, 36, and it is found in Arabic literature. Cf. Ikhwan al- 
Safa, Rasa il. II. pp. 56-57; G. Vajda. "Commcntairc Kairouanais." Revue des etudes jnives, CXII 
(1953). 7, 17, and 18. where reference is also made to the same motif in Fakhr al-Din al-Razi; idem. 
CX1II, (1954). p. 49, line 4. 



232 



Arguments from Design 



Gersonides argues that the ultimate cause of the existence of the heavens is one 
and incorporeal. He takes it as an empirical fact that the course of events on earth 
is directed by the cumulative influence of the stars; and as an astronomical fact 
that every star is imbedded in a celestial sphere, which is sustained in circular 
motion by an "intelligence," a being consisting in pure thought. Now, Gersonides 
argues, each intelligence is responsible for the motion of a given sphere and star 
or stars, 109 and each is aware only of the influences emanating from its star. The 
influence of the various stars interact and dovetail, thereby bringing about the 
total harmonious "order" (nimus) of events on earth. It is well known that when- 
ever artisans belonging to "different crafts" cooperate to produce "one single 
artifact," their cooperation is a result of their being "subservient to," and directed 
by, "a single [master] craft." 110 By the same token, the presence of a single 
overall order in the universe, unless "an accident of some sort," must be due to 
a single "intelligence that has knowledge of the [overall] order of existence" and 
that assigns the appropriate operation to each intelligence and, thence, to each 
star. 111 The celestial intelligences, the heavens, and, through them, the rest of 
the universe must be dependent on a single ultimate cause. Furthermore, the 
cause that governs the intelligences, the cause whose unitary, all encompassing 
thought embraces and subsumes their partial thoughts, must stand to them in the 
relation of "form" to "matter." It must therefore be judged an incorporeal being 
even more certainly than they are judged incorporeal. 1 12 The design evinced by 
the universe thus establishes the existence of a single incorporeal cause of the 
universe. And Gersonides has carried teleological argumentation farther than 
others by formulating what can stand as a complete teleological proof of the 
existence of God. 113 In addition, Gersonides expands his teleological proof of 
the existence of God into a proof of creation; but the reasoning he employs is 
not specifically teleological, and hence not pertinent here. 1 14 

Gersonides' teleological argument for the existence of God was copied by a 
later Jewish writer, Simon Duran (1361-1444). 1 15 Besides repeating Gersonides' 
argument, Duran advances another, much simpler teleological argument based 
on two propositions, the proposition that "the world together with its parts*is 
adapted (na'ot) for the existence of man and for the existence of other beings"; 
and the proposition that "everything adapted . . . and ordered to a single end 

l09 The eighth sphere contains all the fixed stars; by contrast, each of the wandering stars, or 
planets, has its own sphere. 

""The notion is both Platonic and Aristotelian; cf. J. Burnet, The Ethics of Aristotle (London, 
1900), p. xxiv. 

"'Milhamot ha-Shem, V, iii, 8, pp. 272-273. Averroes had offered a similar argument; see 
Epitome of Metaphysics, IV, §§37-38. 
" 2 Ibid., 8, 272-273; 12, 280-283. 

" 3 It should be noted that Gersonides does not advance all the parts of the argument in the same 
place, and that I have brought them together. 

" 4 See above, pp. 209-211. "*Magen Ahot (Livorno, 1785), p. 95a. 



Arguments from Design 



233 



necessarily has a maker." The conclusion Duran immediately draws is that "the 
world has a maker." 1 16 

The foregoing have been teleological arguments for the existence of a cause 
of the universe and, in several instances, for the unity of the cause of the universe. 
In one instance, in Gersonides, the teleological argument for existence and unity 
supplies grounds for the incorporcality of the cause of the universe as well, and 
a complete proof of the existence of God is offered. Arabic philosophic literature 
also knows of teleological arguments that take the existence of God as given and 
attempt merely to establish the presence in the deity of certain attributes, namely 
wisdom or knowledge, power, and providence. Teleological arguments for divine 
attributes usually focus their attention on the details of nature, but they sometimes 
look al nature as a whole. The evidence of design which they cite is either the 
functionality or the marvelous order in nature. The arguments arc especially 
frequent in Kalam works, both in works of Ash'arite writers, who maintained 
that attributes somehow exist as distinct real things in God, and of Mu'tazilite 
writers, who denied that divine attributes l.ave distinct objective existence. The 
K. al-Dala ' il — and by means of it, the tradition going back to the Church Fathers, 
the Stoics, and Socrates — could here too have served as the model. Readers of 
the K. al-Dala il are exhorted to "cogitate upon" and "contemplate" numerous 
features of the celestial and terrestrial regions; for through contemplating the 
functionality of the details of nature, they will be led to recognize God's "gov- 
ernance," or providence, and his "wisdom." 117 

As was mentioned earlier, the K. al-Dala' il is in some manuscripts ascribed 
to the well-known litterateur Jahiz, but the ascription is probably incorrect. The 
genuine Jahiz, a Mu'tazilite, happens to be one of the authors who employed the 
teleological motif. Jahiz assures readers of his work on the animals that what 
prompted him to investigate the seemingly lowly subject was not any intrinsic 
interest the brute animals might have, but instead the evidence they exhibit of 
the " wondrousness of God's governance and the subtlety of his wisdom." 118 The 
thought that design in nature reveals God's wisdom or knowledge plays a role in 
Baqillanl. After having proved the creation of the world and the existence of 
God, Baqillanl asserts that in the domain of human activity, "acts" can have an 
"arrangement and order" only if they come from a "knowing" agent. The prod- 
ucts of the arts of "goldsmithery, carpentry, writing, and weaving" arc examples; 
for in each case, the product undoubtedly entails a craftsman endowed with 
knowledge. Since "God's acts" are far "more subtle and better adapted (ahkamY 
than any human act, they "prove still more plainly" than human acts that the 
agent responsible for them — that is to say, God — is "knowing"; and God must 



" 6 Ibid., p. 4a. 

II7 A'. al-Dalii' il. pp. 2, 5, el passim. 
"*K. al-llayawan (Cairo, 1938). II. p. 109. 



( 



( 

234 Arguments from Design 

I 

( consequently be understood to possess the attribute of "knowledge." 1 19 Similar 

reasoning in the same context recurs in the Mu'tazilite 'Abd al-Jabbar, who, 

' however, unlike BaqillanI denies that divine knowledge has any sort of distinct 

. objective existence. After completing his demonstration of the existence of God 

from creation, 'Abd al-Jabbar writes: "The capability of [executing] a well adapted 

< act" is "proof" that the agent is "knowing." But God's "creation of the animals, 

with the marvels they contain, his producing the circular motion of the spheres, 
his fitting the spheres within one another, 120 his subjugation of the winds, 121 his 

( measuring off of winter and summer — these exhibit finely adapted action to a 

more patent and ample degree than fine writing does." The agent capable of them 
must therefore be deemed "knowing" with more reason than the human agent. 122 

(' Juwaynl, an Ash'arite, advances a teleological argument for the divine attribute 

of "knowledge," again as a corollary to the proof for the existence of God from 
creation. Given the creation of the world and the existence of a creator, Juwaynl 

( submits, "the subtleties of creation, ... the harmony, arrangement, perfection, 

and consummate execution ... of [everything in] the heavens and earth," reveal 
that the agent responsible "is knowing" and in possession of the attribute "knowl- 

( edge." 123 Ghazali's Kalam writings adhere to the same pattern. As a corollary to 

his proof of the existence of a creator, Ghazali finds that the world is "well adapted 
and ordered, perfected and well arranged, encompassing divers marvels." It must 

( accordingly be the handiwork of an agent who is "powerful" and "knowing"; 

and the creator must possess the attribute "knowledge" and the attribute "power." 
To support his conclusion, Ghazali cites the analogy of the art of writing: "Anyone 

( who saw regular lines of writing proceeding in an orderly fashion from a scribe, 

yet who doubted that the scribe has knowledge of the art of writing, would be a 

' fool." Surely, anyone who doubts that the maker of the world has power and 

( knowledge is no less a fool. Ghazali goes further. He calls the conclusion that 

the maker of the world is powerful and knowing a "necessary" inference, by 

' which he means an inference requiring no demonstration because the "intellect 

( confirms it without proof." 124 

In another work, the Ihya, where the stress and tone are devotional, not doc- 

( trinal, Ghazali adduces numerous details from the higher and lower realms, very 

much in the style of the K. al-Dala' il. i2S His subject is the religious virtue of 
"cogitation" (tafakkur), a term that reverberated through the K. al-Dald'il. He 



" 9 K. al-Tamhid (Beirut, 1957), pp. 26, 197. The arguments of Baqillani and the others which 
establish creation and a creator, and to which the arguments for divine knowledge and power are 
attached as corollaries, were discussed earlier in Chapters V and VI. 

l20 Cf. Ibn Hazm's argument, above, p. 226. l2l Cf. Quran 2:164; 45:5. 

' 22 Sharh al-Usul (Cairo, 1965), pp. 156-157. For the nonrcal status of divine attributes, see 
ibid., pp. 182-183. 

'"K. al-Irshad (Cairo, 1950), pp. 61-62; K. al-Shamil (Alexandria, 1969), p. 621 . 
ni al-Iqtisad p al-I'tiqad (Ankara, 1962), pp. 80-81, 99-100; al-Risala al-Qudsiya, Arabic 
section, p. 20; English translation, pp. 40-41. Cf. above, pp. 154-155, 226. 
'"ihya •Ulum al-Din (Cairo, 1937), XV, pp. 78-100. 



( 



Arguments from Design 



235 



bids his reader to "cogitate" over the functionality, order, and "wonders" of the 
details of nature as well as the splendid panorama presented by nature as a whole. 
The reader will thereby be brought to an understanding of God's "knowledge and 
wisdom, the efficacy of his will, his power," "his majesty, and his glory," all with 
the ultimate goal of serving God properly. 126 

Additional illustrations can be given. There is a book entitled al-Hikma fi 
Makhluqat Allah, The Wisdom in God's Creatures, which is ascribed, probably 
in error, to Ghazali. The book is composed of excerpts from the K. al-Dala' il 
and from the devotional work of Ghazali which was just quoted. 127 In it, the 
unknown author underlines the functionality of numerous details of nature. And 
each detail furnishes him with evidence of God's "knowledge," his "governance" 
or providence, his "glory and power, ... the efficacy of his will, and ... his 
wisdom." 128 Bahya, in a devotional context, borrows material from the K. al- 
Dala'il 129 and adds related material of his own, his topic being an exposition of 
the virtue of "reflection" (i'tibcir). The term reflection, as will be recalled, occurs 
in the title of the various recensions of the K. al-Dala' il; 130 and Bahya 's virtue 
of reflection turns out to be very close to the virtue of cogitation which had been 
treated by Ghazali, as reflection is treated by Bahya, in a devotional context. 
"Reflection," writes Bahya, consists in a "contemplation of the signs of divine 
wisdom" in the world. 131 Its object is to bring man to an understanding of God's 
essence and, ultimately, to inculcate proper worship of God. 132 Joseph al-Basir, 
a Jewish Karaite author who usually echoes Mutazilitc thinking, supplements his 
proof of the existence of God with a teleological argument showing God to be 
"knowing." 133 ShahrastanI, in a Kalam work, records teleological arguments for 
the divine attributes of knowledge, wisdom, and power. 134 Judah Hallcvi offers 
a teleological argument for God's "wisdom" and "providence." 135 Ibn Tufayl, in 
no sense a member of the Kalam school, supplements his proof of the existence 
of God with a teleological argument for God's "perfection," "power," "wisdom," 
and "knowledge." 136 Fakhr al-DIn al-Razi 137 and Ijl 131i follow their Kalam fore- 
bears and append to their proofs of the existence of God a teleological argument 
showing the creator to be "knowing," and hence possessed of the attribute of 
knowledge. 



'"'Ibid., pp. 77-78. 

I27 CI\ Baneth, "The Common Teleological Source," pp. 28-29. 
"*al-Hikmafi Makhluqat Allah, pp. 2, 3, 4. 5, 6, 8, 62, 64. 
l2 ''Cf. Baneth. "The Common Teleological Source," pp. 28-29. 

""Above, p. 219. "'al-Hiddya (Hahol ha-Lehahol), II, 1. 

" 2 Ibid., II. introduction. 

L,, P. Frankl, Ein Mu'lazililischer Kalam aus dem 10. Jahrhundert (Vienna. 1872). p. 22. 

1,4 A'. Nihaya al-lqdam, cd. A. Guillaumc (Oxford and London, 1934), pp. 171, 174, 400-401. 

™Kuzari, III. 11; 17; V, 20(1). 

" ( 'Hayy hen Yaqdhan, cd. and trans. L. Gauthier (Beirut, 1936), pp. 88-89; French translation, 
pp. 66-67. 

'"K. al-Arha-in (Hyderabad. 1934), p. 133. lM Mawaqif (Cairo. 1907), VIII. p. 65. 



236 



Arguments from Design 



3. Summary 

Much, though probably not all, Islamic and Jewish argumentation from design 
is a direct outgrowth of a Greek tradition running from Socrates to the Stoics and 
Church Fathers, and thence into Arabic. Islamic and Jewish arguments from 
design, like the ancient arguments, search for design on either a minor or a cosmic 
scale, and as evidence they cite either the functionality or the orderliness and 
beauty of nature. From evidence of design they reason to an intelligent cause of 
the universe; and in some instances they reason from what they discern as a single 
overall plan to a single intelligent planner. A concomitant series of arguments 
from design takes the existence of God as given and contends that the deity, 
however he be known to exist, plainly exercises providence and possesses such 
attributes as wisdom and power. 

Medieval Islamic and Jewish teleological arguments, like the ancient argu- 
ments, generally do not inquire into what constitutes evidence of design; they do 
not justify the step from design to a designer; and they do not explain why a 
single overall plan cannot be a cooperative enterprise by a group— we would say 
a committee — of architects. The stated or unstated presupposition is that no 
intelligent person encountering design will fail to recognize it and that no person 
of sound mind can doubt that design entails an intelligent designer. The best 
thought-out teleological argument would appear to be that of Gcrsonidcs', a 
philosopher who, perhaps significantly, stands not in the Kalam, but in the Aris- 
totelian, tradition. Gcrsonidcs furnishes an Aristotelian explanation of the step 
from design to a designer, and with the aid — unhappily — of broad astrological 
assumptions he is able to establish that the cause of the overall design in the 
universe must be a single agent and incorporeal. The remaining medieval Islamic 
and Jewish arguments seem to be aptly covered by Averroes' evaluation of the 
teleological argument for the existence of God. Averroes characterizes the argu- 
ment as appropriate for instructing nonphilosophers, but as unable to meet the 
standards of formal philosophy. 



VIII 

The Proof from Motion 



1. Aristotle's proof from motion 

Aristotle's proof for the existence of a first mover is not the most widely 
adduced proof of the existence of God in medieval Islamic and Jewish philos- 
ophy, 1 but it does appear frequently. It is to be found in one form or another in 
a work attributed to Alexander of Aphrodisias but known only in the Arabic, 2 in 
the Jabir corpus, 3 in Miskawayh, 4 in a brief work mistakenly attributed to Alfar- 
abi, 5 in Avicenna, 6 Shahrastanl, 7 Ibn Tufayl, 8 Abraham ibn Daud, 9 Averroes, 10 
Maimonidcs," and Aaron ben Elijah. 12 Averroes considered it to be the only 
completely cogent proof of the existence of God. 13 In the same spirit, Maimonidcs 

'The most widely adduced proofs arc discussed in Chapters VI and IX. 

2 Alexander of Aphrodisias, Mabddi' al-Kull. in Arista 'ind al-'Arab, cd. A. Badawi (Cairo. 
1947), pp. 259-263. 

'Jabir ibn Hayyan, Textes Choisis, ed. P. Kraus (Cairo, 1935), pp. 518-521. 

"Miskawayh, al-Fawz al-Asghar, 1, §§3 and 4. The book was not available to mc and I am relying 
on Kh. Abdul Hamid, Ibn Miskawaih. A Study of His al-Fauz al-Asghar, (Lahore. 1946), pp. 14- 
21. 

5 'Uyun al-Masd'il, §13, in Alfarabi's philosophische Abhandlungen, ed. F. Dietcrici (Leiden, 
1890); German translation: Alfarabi's philosophische Abhandlungen aus dem Arubischen uebersetzt, 
trans. I- Dietcrici (Leiden, 1892). p. KH). 

''Avicenna, Nujat (Cairo, 1938), pp. 128-130. 

'Shahrastanl, K. al-Milal wa-l-Nihal, cd. W. Curcton (London, 1846), pp. 385-386 (Shahras- 
tani's account of Aviccnna's philosophy). German translation with pagination of Arabic indicated: 
Religionspartheicn und Philosophenschulen, trans. T. Haarbrueckcr (Halle, 1850-1851). 

8 Ibn Tufayl, Hayy ben Yaqdhan, cd. and trans. L. Gauthier (Beirut, 1936), Arabic text, pp. 84- 
85; French translation, p. 64; English translation: Hayy ibn Yaqzan, trans. L. Goodman (New York, 
1972), p. 132. Ibn Tufayl maintains that a first cause of motion is ipso facto a lirst cause of existence. 

'Abraham ibn Daud, Emuna Rama (Frankfurt, 1852), I, 5; II, i. 

'"Long Commentary on Metaphysics, XII, comm. 41; Arabic text: Tafsir ma hu'd al-Tabi'a, 
cd. M. Bouyges (Beirut, 1938-1948), p. 1632; Epitome of Metaphysics, cd. and trans. C. Quir6s 
Rodriguez (Madrid, 1919), IV, §3; German translation: Die Epitome der Melaphysik des Averroes, 
trans. S. van den Bergh (Leiden, 1924), pp. 105-106. 

"Maimonidcs, Guide to the Perplexed, II, I (I). 

l2 Aaron ben Elijah, 'Es Hayyim, chap. 5. 

"Averroes, Epitome of Metaphysics, IV, §3; German translation, pp. 105-106. 



237 



238 



Proof from Motion 



called it the "strongest argument" and Aquinas called it the "more manifest way" 
of proving the existence of God. 14 

Philosophers and historians have pointed out that two philosophic principles 
lie at the heart of the proof from motion: the principle of causality, and the 
impossibility of an infinite regress of causes. 15 In Aristotle's theory of motion 
the principle of causality expresses itself as the rule that an object can move only 
as long as something sustains its motion. 16 The two philosophic principles just 
mentioned accordingly give rise to an argument to the effect that each moving 
object 17 has a cause sustaining it in motion; the scries of such causes cannot 
regress indefinitely; therefore, the motion of each moving object must be sus- 
tained, ultimately, by a first cause. Here and in other cosmological proofs, first 
cause does not of course mean temporally first cause. Should causes and effects 
happen to precede each other back through time, a first cause would indeed be 
temporally first, as well as first in a more significant sense. But when the links 
in a causal series do not precede each other temporally, the first cause — and this 
is the more significant sense of the term, the sense intended in cosmological 
proofs — is that which, although existing together with the other links, stands 
behind them all and is responsible for the causation running through the series."* 
The necessity of reaching a first cause in the nontemporal sense is what is affirmed 
by the principle of the impossibility of an infinite causal regress. 

In its skeletal formulation, the proof from motion approaches another dem- 
onstration of the existence of God, which 1 shall call the proof from the impos- 
sibility of an infinite regress of efficient causes. 19 That too rests on the two key 
principles, the principle of causality and the impossibility of an infinite regress 
of causes, whence the existence of a first uncaused efficient cause is now inferred. 
The proof from motion and the proof from the impossibility of an infinite regress 
of efficient causes differ most obviously and most fundamentally in that the 
former focuses on causes of motion, whereas the latter abstracts from various 
categories of causation and considers efficient causation in general . The former, 
consequently, arrives at a first mover, which it identifies as the deity, whereas 
the latter arrives at a first efficient cause, which it too identifies as the deity. The 
differing focuses of the proofs will be taken up more fully in a later chapter. 2 " 

Apart from their differing focuses, the proof from motion and the proof from 
the impossibility of an infinite regress of efficient causes also differ in the ways 

l4 Maimonides, Guide, I, 70 (end); Aquinas, Summa Theologiae, I, 2, art. 3, rcsp. (1); cf. 
Aquinas, Commentary on Physics, VIII, §970. 

"Cf. Aquinas, Summa contra Gentiles, I, 13; C. Baeumkcr, Witelo (Muenster, 1908). p. 324; 
H. Wolfson, Philosophy of Spinoza (Cambridge, Mass., 1934), I, p. 193. 

l6 Cf. Aristotle, Physics VII, 2; VIII. 10. 

l7 Motion, here, means not merely locomotion, but change in general. Cf. Physics VIII, 7, and 
below, n. 33. 

'"This is the second sense of priority defined by Aristotle in Categories, 12. 
l9 Cf. below, Chapter XI. 
20 Below, pp. 337, 344. 



Proof from Motion 



239 



they elaborate the central, skeletal argument. The reason why the skeletal argu- 
ment must in each instance be elaborated is that a first cause is not necessarily a 
deity. The first cause reached by tracing back a scries of moving causes or efficient 
causes might be an inanimate object; it might be an animate physical object; or 
there might exist a multitude of first causes. Until the incorporeality and unity of 
the first cause are established, the minimum specifications for a deity, at least in 
the view of the medieval Aristotelian philosophers, are not met. 21 Both the proof 
from motion and the proof from the impossibility of an infinite regress of efficient 
causes expend more effort on establishing the incorporeality and unity of the first 
cause than they do on establishing the existence of such a cause. 

Aristotle, the originator of the proof from motion, provided arguments for both 
the incorporeality and the unity of the first mover. His argument for the incor- 
poreality of the first mover took its departure from the eternity of motion; he 
reasoned that since no corporeal object could contain power sufficient to sustain 
eternal motion, the first mover cannot be a corporeal object. 22 Aristotle provided 
two arguments for the unity of the first mover. One inferred the unity of the 
mover from the unity of the fundamental underlying motion of the universe; 23 
the other, which appears in a different work, deduced the attribute of unity from 
the attribute of incorporeality. Here the reasoning was: "All things that are many 
in number have matter"; and since the first mover has no matter, it cannot be 
many in number. 24 Recent scholars have questioned whether Aristotle's most 
considered position was indeed that only a single first mover exists, 25 but medi- 
eval writers had no such doubts. 

Whereas the skeletal argument is simple and direct, Aristotle's complete proof 
of the existence of God from motion is quite complex. The skeletal argument is 
supplemented not only by arguments for incorporeality and unity, but also by an 
argument designed to show that the first cause of motion is, more comprehen- 
sively, the first cause of all types of change occurring in the universe. The com- 
plete proof winds tortuously through Books VII and VIII of Aristotle's Physics, 26 

21 Cf. Maimonides, Guide, I, 71 (end). "Aristotle, Physics VIII, 10. 

"Aristotle, Physics VIII, 6, 259a, 13-20. "Aristotle, Metaphysics XII, 8, 1074a. 33-37. 

25 Cf. H. Wolfson, "The Plurality of Immovable Movers in Aristotle, Avcrrocs. and St. Thomas," 
reprinted in his Studies in the History of Philosophy and Religion,\a\. I (Cambridge. Mass.. 1973), 
pp. 1-21; D. Ross, "The Development of Aristotle's Thought," in Aristotle and Plato in the Mid- 
Fourth Century, ed. I. Dueling and G. Owen (Gocteborg, 1960), pp. 12-14; and, for further bibli- 
ography, D. Frcdc, "Theophrasts Kritik am umbewegten Bewcgcr." Phronesis, XVI (1971). 65. 

26 The relationship of Physics VII \o Physics VIII is problematical. Cf. Averroes, Long Commen- 
tary on Physics, in Arislolelis Opera cum Averrois Commenlariis, Vol. IV (Venice, 1562), VIII, 
comm. 9; idem, Derushim Tih'iyim. ed. H. Tunik, Ph.D. dissertation, Radcliffc College (1956), §7; 
Aquinas, Commentary on Physics, VII, introduction; H. Wolfson, "Notes on Proofs of the Existence 
of God in Jewish Philosophy," in his Studies in the History of Philosophy and Religion, Vol. I, p. 
580; K Solmscn, Aristotle's System of the Physical World (Ithaca, 1960), p. 228, n. 19. All of these 
find that the discussion in Physics VIII relies on the conclusions of Physics VII. D. Ross in the 
introduction to his edition of the Physics (Oxford, 1936), pp. 15-19, finds that Book VII is an 
interpolation into the core of the Physics. 



240 



Proof from Motion 



and many more than the two central philosophic principles come into play. In the 
Middle Ages the full argumentation of Aristotle's proof was unraveled, reassem- 
bled, and presented afresh by Averroes, 27 Maimonides, and Aquinas. 28 Most 
systematic is Maimonides' version. It painstakingly spells out the principles and 
steps, implied or explicit, the full proof being shown to rest on no less than 
nineteen formal principles of Aristotelian philosophy. Maimonides' articulation 
of the premises and steps in the proof facilitated and invited critical scrutiny; and 
his version, did, in fact, elicit a most searching critique by Hasdai Crescas. 
Crescas does not transcend his age. He does not reject, in totality, the overall 
Aristotelian framework within which the proof is worked out. And he does not 
question the proposition in the proof which would be most questionable for the 
modern reader, to wit, the principle that motion continues only as long as sus- 
tained by a cause. 29 Crescas' critique is, however, as radical a critique as could 
be drawn up in Aristotelian terms by a student of Aristotelian physics. 30 

The present chapter will examine Maimonides' version of the proof from motion, 
that version being considered as a systematic articulation of what is explicit and 
implied in the proof. Then Crescas' critique will be examined, it being considered 
as a capital illustration of the extent to which the proof can be subverted from 
within the framework of Aristotelian physics. Finally, an offshoot of the proof 
from motion will also be examined together, again, with Crescas' critique. 

2. Maimonides' version of the proof from motion 

The proof from motion as reformulated by Maimonides carefully establishes 
each of the theses that are, in his view and in the view of other medieval philos- 
ophers, required for a complete demonstration of the existence of God. That is 
to say, he undertakes to establish (a) the existence of a first cause, (b) the incor- 
poreality of that cause, and (c) its unity. In the course of establishing the three 
theses, Maimonides, as already mentioned, cites nineteen principles of Aristo- 
telian philosophy. Each of those principles has its philosophic justification in 
Aristotle, but Maimonides states them flatly, without supporting reasoning, because 
he was wary of overwhelming his reader with technical detail. 31 Crescas in his 
critique will, before setting forth his own objections, lay bare the presuppositions 
and reconstruct the reasoning underlying each of the principles. 

( a) The existence of a first mover 

The object of the Aristotelian proof from motion is to establish a first cause of 
locomotion, or motion in place. In addition, though, the proof proposes to trace 



"See above, n. 10. Summa conlra Gentiles, I, 13. 

29 That principle stands in opposition to Newton's first law of motion, the law of inertia. Cf. A. 
Kenny, The Five Ways (London, 1969), pp. 28-31. 

30 Cf. H. Wolfson, Crescas' Critique of Aristotle (Cambridge, Mass., 1929), pp. 114-127, 
especially pp. 125-127. 

"Maimonides, Guide, I, 71. 



Proof from Motion 



241 



all other changes in the universe to motion in place, so that the first cause of 
motion in place will emerge, more comprehensively, as the first cause of all 
change and motion in the universe. 32 Maimonides in his systcmatization of the 
proof accordingly cites principle (1) affirming that there are "four categories" in 
which change occurs, namely "the category of substance, ... of quantity, . . . 
of quality, . . . and of place." 33 In reasoning back to a first cause of motion in 
place, Maimonides starts by considering a single given instance of change in the 
category of substance: He inspects the coming into existence of an actual physical 
object. Principle (2) states that every "individual compound substance" has "mat- 
ter and form" as its components, and also that a substance of the sort comes into 
existence only after some factor "prepares" the given matter to receive the appro- 
priate form. 34 Here, as will be observed, is the principle of causality applied to 
the most radical type of change, the coming into existence of a new object. 

Whenever a given portion of matter receives a new form, some factor, then, 
prepares the matter. If the factor preparing a portion of matter to receive a new 
form should itself be brought to the state wherein it performs its causal function, 
a further causal factor must be responsible for the change; here the principle of 
causation is applied to changes of state. If that factor, in turn, is also brought to 
the state wherein it performs its causal function, there must lie behind it yet 
another factor. According to principle (3), however, an infinite regress of causes 
is impossible. 35 The scries of causal factors lying behind the generation of a given 
physical object cannot, therefore, regress indefinitely, and a first term in the scries 
must be posited. 36 Now, whereas the end product of the series that Maimonides 
is considering is a change in substance, the change directly brought about by the 
first cause in the series must be motion in place. For principle (4) affirms that 
"motion in place" is the "most primary" of the four categories of change, inas- 
much as a change of place ultimately stands behind all the other kinds of change. 37 
The series of causes leading up to each change in the universe has thus been 
traced back to a first cause whose immediate effect is motion in place. 

The generation of a single given substance is not an isolated event; it is part 
of a continuing process. And for the processes of nature to continue uniformly, 
without interruption, the underlying motion in place from which other motions 



"Cf. Physics VIII, 7. 

"Guiile. II, introduction, prop. 4. Cf. Physics III, 1, 201a, 4-9; V, 1, 225a, 34 ff. In the former 
passage Aristotle states that "motion" and. "change" take place in the four categories. In the latter 
passage, he observes that motion and change take place in the categories of quantity, quality, and 
place; but in the category of substance only change, not motion, occurs. 

'"Guide, II, introduction, prop. 25. 

"Ibid., prop. 3. Cf. Aristotle, Physics VIII, 5, 256a, 4— 256b, 3; Metaphysics II, 2, 994a. 5- 
8; 1 1-19; below, p. 337. 

36 The first cause is first in the sense defined above, p. 238. It is also temporally prior to at least 
some effects in the scries, namely those effects that come into existence after not having existed. 

31 Guide, II, introduction, prop. 14; Aristotle, Physics VIII, 7. 



242 



Proof from Motion 



and changes derive must be continuous. 38 Principle (5) states that circular motion 
in place is the sole type of change capable of continuing indefinitely and without 
interruption. 39 The effect directly produced by the first cause of a given event — 
that is to say, by the first cause of the continuing process of which the given event 
is a part — must hence specifically be circular motion in place. Side by side with 
these abstract considerations, the analysis of actual events in the sublunar world 
revealed to the Aristotelian observer that all sublunar events are traceable to the 
circular motion of the heavens. 40 Maimonides consequently affirms that every 
motion and change in the sublunar world is traceable to the continual circular 
movement of the heavens, and that the cause of the motion of the heavens is the 
ultimate cause of all motion and change in the sublunar world. 

To complete the argument and raise it to a genuine demonstration of the exis- 
tence of God, the first cause of the movement of the heavens must be shown to 
be an incorporeal being beyond the heavens, and there must be shown to exist 
only one first incorporeal cause. Restating the same in the terms of Aristotelian 
astronomy, the cause of the motion of the celestial spheres must be shown to be 
a single incorporeal being beyond the spheres. Once that has been proven, a 
single incorporeal first cause of all motion and change in the universe will have 
been established. 

(b) The incorporeality of the first mover 

Maimonides hereupon cites principle (6), which is, again, the principle of 
causality in one of its guises. Principle (6) states that "whatever undergoes motion, 
necessarily has a mover," that is to say, a cause sustaining motion as long as the 
object continues to move; 41 and, the same principle continues, the cause of motion 
obviously exists either "outside" the moving object or "within it." 42 The system 
of celestial spheres must hence be sustained in motion either by something out- 
side, or else by something within, the system of spheres. Each of those two 
possibilities, the possibility that the mover of the spheres is to be found outside 
the spheres and the possibility that the mover is within, subdivides in two; and 
a total of four alternatives results. For should, on the one hand, the mover of the 
spheres lie outside them, the mover must be either — the first alternative — another 
body, or — a second alternative — an incorporeal entity. And should, on the other 
hand, the mover lie within, principle (7) comes into play. Principle (7) states that 
whatever is in a body is a power of one of two conceivable sorts. It is either a 
power that strictly does "exist within the body," or else it is a power "through" 
which "the body exists, such as the natural form." 43 Accordingly, should the 



38 This is not stated explicitly by Maimonides, but cf. Aristotle, Physics VIII, 6. On the problem- 
atic character of the proposition, sec Ross's introduction to his edition of the Physics, pp. 91-92. 
,9 Guide, II, introduction, prop. 13; cf. Aristotle, Physics VIII, 8. 

'"Guide, II, 1 (1). Cf. Aristotle, Metaphysics XII, 6, 1072a, 10-18; De Generatione el Corrup- 
lione II, 10. 

"'Cf. Aristotle, Physics VII, I; VIII, 10. 

41 Guide, II, introduction, prop. 17. "Guide, II, introduction, prop. 10. 



Proof from Motion 



243 



mover of the spheres be present within the body of the spheres, it would be 
cither — a third alternative — strictly within the body of the spheres, in other 
words, a physical "power distributed through the entire sphere [or spheres)" and 
hence "divisible"; 44 or else — -a fourth alternative — the mover would be a natural 
form that is "not divisible," in olhcr words, a "soul" or "intellect." Intellect in 
this fourth alternative is to be distinguished from incorporeal entity in the second 
alternative. Intellect here has the sense of a soul, possessing the faculty of reason, 
which is attached to a body. The incorporeal entity in the second alternative is 
an entity consisting in pure thought, that is to say, an intellect, which is not 
attached to a body. 

The conceivable ways of construing the cause of celestial motion are thus 
exhausted by four alternatives. For the purpose of his exposition, Maimonides 
arranges the four alternatives in the following order. The cause of the motion of 
the spheres and, thereby, the ultimate cause of motion and change in the universe 
must be either: (i) a body beyond the spheres; (ii) a power distributed through 
the body of the spheres — or, put more generally, a power distributed through the 
body of the universe; (iii) a natural form or "power" present in, but not distributed 
through, the. body of the spheres, in effect, a "soul" or "intellect" attached to 
the spheres; or (iv) a purely incorporeal entity, distinct from the spheres. Mai- 
monides undertakes to establish the correctness of alternative (iv) by eliminating 
the other three alternatives. In formulating four exhaustive alternatives and elim- 
inating three of them in order to establish the correctness of the fourth, Mai- 
monides is introducing something not found in Aristotle. There he is systematiz- 
ing the proof. However, the overall contention that the unceasing movement of 
the spheres implies an incorporeal mover is genuinely Aristotelian. 45 And all of 
the argumentation whereby Maimonides eliminates the three unacceptable alter- 
natives is drawn from Aristotle's statement of the proof. Maimonides seems even 
to have taken pains to find a role, a niche, for strands of argumentation in Aristotle 
which he could well have regarded as superfluous or redundant. 

(i) The alternative that the mover of the spheres is a boa 1 )! 
beyond the spheres. 

If the system of spheres should be moved by a body beyond the spheres, 
Maimonides reasons, the body in question must itself undergo motion because 
of principle (8); principle (8) states that one body can move another only when 
it is itself in motion. 46 The body moving the spheres would likewise have to have 
a cause of its motion. And if it were moved by yet another body, the latter too 
would have to undergo motion and would require a cause of its motion. To assume 
an infinite regress of moving bodies would entail the simultaneous existence of 
an infinite number of bodies. But the simultaneous existence of an infinite number 
of bodies is ruled out by principle (9), which affirms, in general, the impossibility 

44 Cf. ibid., prop. II. 45 Cf. Physics VIII, 10. 

^Guide, II, introduction, prop. 9; cf. Aristotle, Physics VII, 1. 



244 



Proof from Motion 



of the simultaneous existence of an infinite number of magnitudes. 47 Conse- 
quently, a cause of the motion of the spheres must eventually be reached which 
is not a body. 

It should be noted that principle (9) and the inference drawn therefrom are 
superfluous. Maimonides had already established a first cause of motion with the 
aid of principle (3), the impossibility of an infinite regress of causes. He could, 
therefore, now more appropriately and more simply have contended that the first 
cause of motion already established cannot be a body; for it would then be in 
motion, hence have a cause of its motion, and not after all be the first cause. 
Something undoubtedly led Maimonides to rule out an infinite regress specifi- 
cally of moving bodies, in so curiously oblique a fashion — by arguing that such 
a regress would entail the assumption of an infinite number of bodies, whereas 
the existence of an infinite number of bodies is impossible. What did lead Mai- 
monides to argue as he does can be surmised. A similar argument appears in 
Aristotle's Physics VII, 48 and Maimonides presumably wished to find a niche for 
that argument of Aristotle's in his own restatement of the proof from motion. 

In any event, the system of celestial spheres and, in fact, every scries of moving 
bodies, has been shown to owe its motion to a cause which is not a body. The 
question recurs: What is the nature of the cause? 

Maimonides proceeds to adduce an argument showing that the first cause of 
motion cannot be a power distributed through a body, and a separate argument 
showing that the cause cannot be a power of the type that, although not distributed 
through a body, is present in one. 

( ii) The alternative that the mover of the spheres is a power 
distributed through a body 

Aristotle had established that the first mover cannot "have magnitude" through 
an argument running thus: Every entity having magnitude, including the corporeal 
universe as a whole, must be finite, since nothing of infinite magnitude can 
possibly exist. Whatever is of finite magnitude can contain only finite power. 
Finite power can produce motion for only a finite time. But motion, specifically 
the motion of the celestial spheres, is eternal and therefore dependent upon an 
infinite power. From these premises, Aristotle concluded that the first mover must 
be "indivisible, without parts, and without magnitude." 49 It must, in other words, 
be an incorporeal being. Without spelling out the several alternatives to an incor- 
poreal first mover, Aristotle excluded them all through a single argument. 

"Guide, II, introduction, prop. 2. 

^Physics VII, 1 ; here Aristotle does not establish that an infinite number of bodies cannot exist, 
but rather than an infinite number of bodies would be unable to move. 

49 Physics VIII, 10, summed up in 267b, 17-26. The argument has four premises: every body is 
finite; a finite body can contain only finite power; a finite power can produce motion for only a finite 
time; the motion of the heavens is eternal. In Maimonides' reformulation, the principle that a finite 
power can produce motion for only a finite time will not have the status of a separate premise. 



Proof from Motion 



245 



As Maimonides read Aristotle, he apparently found the argument just outlined 
to be inappropriate for eliminating alternative (i), the alternative that the mover 
of the spheres is another body. That alternative, as just seen, is eliminated by 
Maimonides through a different argument. 5 " And Maimonides apparently found 
Aristotle's argument inadequate for eliminating alternative (iii), the alternative 
that the mover of the spheres is a power of the type which is not distributed 
through a body. While one can only guess why Maimonides considered the 
argument inappropriate for eliminating alternative (i), 51 a reason why he might 
find it inadequate for eliminating alternative (iii) is easily discovered: A power 
not distributed through a body is not strictly contained in a magnitude and con- 
sequently is not covered by the consideration that a finite magnitude can contain 
only finite power. 52 The single purpose for which Maimonides docs use Aristotle's 
argument is to eliminate alternative (ii), the alternative that the spheres are moved 
by a power of the type which is distributed through a body. 

In reformulating Aristotle's argument, Maimonides cites principle (10), according 
to which every power distributed through a body is divisible in the same manner 
that its body is divisible. 53 Principle (11) adds that no body can be "of infinite 
magnitude." 54 Together, (10) and (1 1) imply (12), the principle that every power 
distributed through a body is finite. 55 A finite power could not, however, produce 

"'Above, pp. 243-244. 

"The argument strictly eliminates the thesis that the mover of the spheres is a power distributed 
through a body, not the thesis that the mover of the spheres is a body; but I am not sure that Maimonides 
could conceive of a body's producing motion by itself, i.e., by its mass as distinct from its power. 
In any event, though. Maimonides would be led to use a separate argument for eliminating alternative 
(i) by his desire to utilize every strand of argumentation provided by Aristotle; cf. above, p. 244. 

52 Cf. S. Munk's note to his translation of Maimonides, Le Guide des Egares. Vol. II (Paris, 
1861), p. 32, n. 2. 

"Guide. II, introduction, prop. II. 54 lbid., prop. 1; cf. Physics III, 5. 

"Guide. II, introduction, prop. 12. The reasoning whereby (10) and (11) imply (12) would be 
as follows: The power distributed through a body, being divisible, may be assumed actually to be 
divided. Since the body is finite, it can only be divided into a finite number of parts, and the power 
distributed through it, which is similarly divided, will also have a finite number of parts. Each partial 
power will, moreover, clearly be finite; for each part will be exceeded by the larger whole, and 
nothing that is exceeded can be infinite (cf. above, p. 88). Inasmuch as each partial power is finite 
and the number of partial powers is finite, the total power must be finite. 

This train of reasoning — which is similar to the argument whereby Aristotle proves that a finite 
power can produce motion for only a finite time, Physics VIII, 10, 266a, 12 ff. — is spelled out by 
Alexander, cited in Simplicius, Commentary on Physics, cd. H. Diels. Commenlaria in Arislotelem 
Graeca. Vol. X (Berlin, 1895), p. 1326; Avicenna, Najat (Cairo, 1938), p. 130; Ghazali. Maqasid 
al-Falasifa (Cairo, n.d.), p. 209; Ibn Tufayl, Hayy hen Yaqdhan. Arabic text, p. 84; French translation: 
p. 64; English translation: p. 84; Shcm Tob's commentary to Maimonides' Guide, II, introduction, 
prop. 12. Also cf. Munk's translation of the Guide, II, p. 32, n. 2. Aristotle proves principle (12) 
differently; cf. Physics VIII, 10, and below, p. 261. In an article entitled "The Principle that a Finite 
Body Can Contain Only Finite Power," Studies in Jewish Religious and Intellectual History Presented 
to Alexander Allmann (University, Alabama, 1979), p. 77, I suggest a reason why Alexander and 
the others might have wished to substitute their reasoning for Aristotle's. 



246 



Proof from Motion 



motion for an infinite time; 56 this proposition is so obvious to Maimonides that 
he does not even register it in his list of formal principles. Yet principle (13), 
which Maimonides accepts merely for the sake of argument, asserts that the 
universe and its motion are eternal. 57 Since eternal motion has to be sustained 
by infinite power, and since no power distributed through a body can be infinite, 
the eternal motion of the spheres, so Maimonides concludes, could not be due 
to a power distributed through a body. 

The alternative that the motion of the spheres is due to a power distributed 
through a body is thus ruled out. Earlier the alternative that the motion of the 
spheres is due to a body beyond the spheres was also ruled out. Two conceivable 
explanations still remain: The cause of the motion of the spheres must be either 
a power present within, but not distributed through, the body of the spheres; or 
else it is an incorporeal being. 

( Hi) The alternative that the mover of the spheres is a power 
present in, but not distributed through, the body of the 
spheres 

In the course of his proof of the existence of a first incorporeal mover, Aristotle 
had at one point made the statement that the first cause of eternal motion cannot 
itself undergo motion accidentally, "since what is accidental . . . has the possi- 
bility of not being." 58 Maimonides, characteristically, utilizes the thought embod- 
ied in Aristotle's statement for a single specific purpose, to show that the ultimate 
mover of the spheres cannot be a power present within, although not distributed 
throughout, the body of the spheres, in other words, to show, that the motion of 
the spheres is not produced by a soul attached to the spheres. 

Maimonides' formulation of the argument begins with principle (14), which 
affirms that physical bodies and nothing else can undergo motion "essentially." 
By contrast, the same principle continues, powers in a body, whether or not they 
are distributed through their body, do undergo motion "accidentally" as a con- 
sequence of their presence "in" something that undergoes motion essentially. 59 
Principle (15) adds that nothing undergoing motion accidentally can continue to 
move indefinitely; 60 for the accidental, having the possibility of not existing, 
cannot be eternal. 61 It follows that every power in a body must eventually come 



56 Guide, II, 1 (1); cf. Aristotle, Physics VIII, 10. 

51 Guide, II, introduction, prop. 26; cf. Physics VIII, 1, and above, Chapters II and III. Mai- 
monides' position is that the proof from motion is needed only on the hypothesis of eternity, whereas 
on the opposite hypothesis the existence, incorporeality, and unity of the creator can be inferred from 
the world's having come into existence. Cf. Guide, I, 71; II, 2; above, pp. 4-5. 

'"Physics VIII, 5, 256b, 10; cf. ibid., 6, 259b, 28-31. 

59 Guide, II, introduction, prop. 6; cf. Aristotle, Physics VIII, 4, 254b, 7 ff. 

m Guide, II, introduction, prop. 8; cf. Aristotle, Physics VIII, 5, 256b, 8-10. 

"'The principle that every possibility must eventually be realized was met above, p. 90, and will 
be met again below, pp. 320, 381. 



Proof from Motion 



247 



to rest. When a power in a body would come to rest, the body inextricably 
associated with it and moved by it would also necessarily come to rest. 62 As 
already seen in principle (13), our discussion is proceeding on the hypothesis 
that the motion of the celestial spheres is eternal. The primary motion in the 
universe cannot, consequently, be due to a power in a body, even a power of the 
type that is attached to, without being distributed throughout, its body; for such 
a power could not sustain eternal motion. 

The reasoning put forward here, as will be observed, applies equally to powers 
that arc, and powers that are not, distributed through a body, as long as those 
powers arc properly described as in a body. Powers of both types would, by the 
reasoning given, undergo motion accidentally, and hence not be able to undergo 
motion indefinitely; and whatever is inextricably attached to them and moved by 
them would likewise be unable to undergo motion indefinitely. Maimonides could 
thus have utilized the present reasoning earlier when he was eliminating the 
previous possible explanation of celestial movement, the alternative that (ii) the 
cause of the motion of the spheres is a power distributed through the spheres. He 
could have dispensed with the argument to the effect that a finite body can contain 
only finite power, and finite power can sustain motion for only a finite time; and 
he could have reasoned instead that a power distributed through a body would 
itself undergo accidental motion and therefore be unable to produce motion eter- 
nally. Maimonides presumably employed the other train of reasoning when ruling 
out alternative (ii) because of his desire to find a role for each strand of argu- 
mentation in Aristotle's proof from motion. 

(iv) The remaining alternative: The cause of the motion of the 
spheres is an incorporeal being 

The cause of the motion of the celestial spheres has now been shown to be 
neither a body, nor a power distributed through a body, nor a power present in, 
although not distributed through, a body. The remaining alternative is that the 
first cause of the motion of the spheres and, thereby, the ultimate cause of motion 
and change in the sublunar world is (iv) an incorporeal being. Principle (16) 

62 This step raises difficulties. Maimonides has argued not merely that a power undergoing acci- 
dental motion cannot produce motion eternally; he has argued that such a power or anything else 
undergoing accidental motion cannot itself be in motion eternally. That would mean that the spheres 
cannot move eternally — cannot even be maintained eternally in motion by a transcendent incorporeal 
cause — should anything attached to them undergo accidental motion. For example, the spheres would 
be unable to move eternally if a soul is inextricably attached to them; or. as Narboni observed in his 
commentary, if such qualities as transparence and radiance arc attached to them; or indeed, as Crcscas 
observed, if their own surface is attached to them! For the soul, the qualities transparence and 
radiance, and the surface of the spheres, arc not bodies. Hence they move not essentially but acci- 
dentally and. by Maimonides' reasoning, would eventually have to cease moving. Cf. Narboni's 
commentary on Guide, II. introduction, prop. 8; Crcscas, Or ha-Shem, I, i, 8; Wolfson. Crescas, 
pp. 250-251. 551-561; and Aristotle. Physics VIII, 6, 259b, 28-31. Below, p. 271, I shall suggest 
that there is an equivocation in Maimonides' use of the term accidental. 



I 



248 Proof from Motion 

states that whatever is capable of undergoing motion is divisible (and, conversely, 
whatever is not capable of undergoing motion is indivisible). 63 Principle (17) 
states that motion is a kind of change (and change is a kind of motion). 64 Since 
the first cause of motion cannot, as already seen, undergo motion cither essentially'' 5 
or accidentally, 66 it must by principles (16) and (17) be indivisible and exempt 
from all change. Principle (18) affirms that "time is an accident consequent upon 
motion. . . . and whatever is exempt from motion does not fall under time." 67 
The unmoved first mover hence exists outside the domain of time. 

The first mover must then be an indivisible, unchanging, timeless, incorporeal 
substance. 

(c) The unity of the first mover 

In Aristotle's fullest statement of the proof from motion, the unity of the first 
mover is inferred from the unity of the underlying primary motion of the universe, 
a single continuing motion implying a single mover. 68 In a condensed statement 
of the proof from motion, in the Metaphysics, Aristotle derived the unity of the 
^ first mover from the attribute of incorporeality. 69 The latter method is employed 

( by Maimonides. 

Principle (19) is a version of the rule that matter is the ground whereby objects 
are individuated. According to principle (19), "what is not a body . . . or a power 
in a body . . . cannot conceivably by enumerated" except in one instance. The 
sole instance in which incorporeal beings can be differentiated from each other 
and enumerated is when they stand in the relation of "causes and effects." 70 
Within the class of causeless beings the differentiation and enumeration of sepa- 
rate entities as causes and effects would of course be impossible; one causeless 
being obviously cannot be the cause or the effect of another causeless being. 
Consequently, no more than a single incorporeal uncaused mover can exist. Yet 
< should there also exist other incorporeal movers of the spheres, in addition to the 

f first unmoved mover, those other incorporeal movers would have to stand in 

j-- relation to each other as causes and effects; for only then would they be distin- 

guishable from each other and numerable. Thus principle (19) arranges the incor- 
f poreal movers in the Aristotelian scheme of the universe into a causal series with 

^ a single uncaused cause at the apex. 

( 

"Guide, II, introduction, prop. 7; cf. Aristotle, Physics VI, 4, 234b, 10-20. In this and the 
following proposition 1 have added the clauses in brackets. Without the additions, the logic of the 
inference Maimonides is about to draw would be faulty. 
M Guide, II, introduction, prop. 5. 
"Alternative (i). 
"Alternatives (ii) and (iii). 
f "Guide, II, introduction, prop. 15; cf. Aristotle, Physics IV, 11. 

L "Physics VIII, 6, 259a, 13-20. 

f 69 Above, n. 24. 

^ 10 Guide, II, introduction, prop. 16. The principle is never stated in just this way by Aristotle. 

\ 



f 



Proof from Motion 



249 



With the unity of the (irst mover, the proof from motion is complete. The proof 
has established the existence of a single incorporeal first mover of the celestial 
spheres and, through the mcdiacy of the spheres, of the sublunar world. 

3. Hasdai Crcscas' critique of the proof from motion 

Crescas' critique of the proof from motion contains at least ten separate objec- 
tions arranged in an artificial scheme. 71 In the present section, I am rearranging 
Crcscas' objections to conform to the three main moments in the proof, that is, 
the arguments for the existence, for the incorporeality, and for the unity of the 
first mover. Crcscas' critique of the arguments for incorporeality and unity are 
of greater interest than his critique of the argument for the existence of the first 
mover. And his critique of the argument for incorporeality is so far ranging that 
the train of thought is difficult to follow. Crescas' intent throughout is not free- 
thinking, but conservative. His aim is to dispute not the existence, incorporeality, 
or unity of God, but the ability of philosophy to demonstrate those three theses. 

(a) The existence of a first mover 

Principle (3), the impossibility of an infinite regress of causes, is essential to 
the proof from motion. It alone allows the causes behind the appearance of a 
given physical object to be traced back to a first term; and unless a first term in 
the series of causes is reached, the proof accomplishes nothing. Crescas ostensibly 
refutes principle (3), the impossibility of an infinite regress of causes, thereby 
ostensibly invalidating the proof entirely. 72 In fact, the principle that causes can- 
not regress infinitely contains two separate points. The principle asserts both that 
every chain of causes must terminate at a first cause and that the number of links 
in a causal chain must be finite. Of the two points, only the former is essential 
for demonstrating the existence of God, whereas, once a first cause is posited, a 
demonstration of the existence of God can dispense with the latter point, the 
proposition that the number of causes and effects must be finite. And Crcscas' 
refutation is addressed exclusively to the latter point. He acknowledges that a 
first cause must indeed be posited for every series of causes and effects, and 
merely contends that, granted a first cause, the number of links between it and 
the final effect might conceivably run to infinity. 73 Crescas' refutation of the 
principle that causes cannot regress infinitely thus in no way goes to the heart of 
the proof from motion. His reservations on the principle are of some interest, 
but they will best be examined in a later chapter; the grounds for the impossibility 
of an infinite regress which were known to him grow out of a certain proof for 

7I Crcscas lists seven principles employed in the proof which, he contends, are unsubstantiated. 
Then he gives three reasons why the proof is invalid even if all the principles are granted. Or ha- 
Shem. I, ii, 15. 

12 Or ha-Shem, I, ii, 3; 15 (following the reading of the Vienna edition); Wolfson, Crescas, pp. 
224-229. 

"Cf. below, p. 365. 



( 

( 
( 

250 Proof from Motion 

( the existence of God advanced by Aviccnna, and his refutation will be best taken 

up after examining Aviccnna's proof. 74 At the same time I shall also take up a 
' more trenchant critique of the principle that causes cannot regress infinitely, a 

I critique developed by Ghazali. 75 

( (b) The incorporeality of the first mover 

( Maimonides established the incorporeality of the first mover by laying down 

the Aristotelian principle (6), according to which objects move only as long as 
( they are sustained in their motion by a cause, and by eliminating all but one of 

j the conceivable ways of construing the cause of celestial motion. The mover of 

the spheres cannot, Maimonides argued, be (i) a body beyond the spheres, (ii) a 
^ power distributed through a body, or (iii) a power present in, although not dis- 

( tributed through, a body, in effect, a soul of the spheres; therefore the mover of 

the spheres must be the sole alternative remaining, (iv) an incorporeal being 



( independent of the spheres. 76 Crescas challenges the conclusion. But whereas a 

( modern critic might question principle (6), which affirms that motion continues 

only as long as sustained by a cause, 77 Crescas has no thought of taking so radical 
( a course. His procedure is instead to refute the several arguments whereby each 

I of the three alternatives to an incorporeal mover had been ruled out. 

( (i) The alternative that the mover of the spheres is a body 

( beyond the spheres 

To rule out the alternative that the mover of the spheres is a body beyond the 
* spheres, Maimonides had reasoned: One body can produce motion in another 

( only if it is itself in motion. To assume that every body producing motion is 

moved by another amounts to assuming an infinite series of moving bodies. And 
( that is excluded by principle (9), the impossibility of the simultaneous existence 

( of an infinite number of magnitudes. The mover of the spheres cannot therefore 

be a body beyond the spheres; or, to be more precise, if the mover of the spheres 

is a body beyond the spheres, a mover must ultimately be reached which is not 
, a body. 78 

Crescas' critique consists in rejecting the various grounds that had been adduced 
( to support principle (9), the impossibility of an infinite number of magnitudes. 

^ Ostensibly, he thereby defends the thesis that motion in the universe may^bc 

explained through an infinite regress of moving bodies all existing together. 79 In 
\ fact, Crescas' refutation of principle (9) does not affect the proof from motion, 

^ 74 Below, ibid. 

( "Below, pp. 366-372. 

76 Cf. above, pp. 243 - 247. 

( "Crescas will propose the thesis that the elements are moved to their proper places, and the 

spheres are moved circularly, by their natures; cf. below, p. 269. But that falls far short of Newton's 
first law of motion. 

( 78 Above, pp. 243-244. 79 0r ha-Shem, I, ii, 2; 15; Wolfson, Crescas, pp. 218-221. 



( 



Proof from Motion 



251 



and for that reason need not be explored here. 80 For even without principle (9), 
an infinite regress of moving bodies is excluded by the more fundamental prin- 
ciple — principle (3) — which affirms the impossibility of an infinite regress of 
causes. Although Crescas rejected part of principle (3) as well, he accepted its 
more significant part, namely, the proposition that every series of causes and 
effects has a first cause; 81 and he thus cannot consistently defend a picture of the 
universe in which each body is moved by another body and a first cause is never 
reached. Crescas was too deeply under the sway of Aristotelian physics to reject 
the proposition that every motion must be sustained by a cause, or the proposition 
that bodies produce motion only when themselves in motion, or again the prop- 
osition that causes cannot regress indefinitely without reaching a first cause. And 
unless one of those propositions is rejected, the Aristotelian conclusion is una- 
voidable that motion is ultimately due to a first cause which is not a body. 

It is in what follows that Crescas genuinely comes to grips with the proof from 
motion. 

(ii) The alternative that the mover of the spheres is a power 
distributed through a body 

Even granting that the Aristotelian proof from motion can establish a first cause 
which is not a body, the proof still has to show that the first cause is not a power 
in a body. Should the first cause of motion be the kind of power which is distrib- 
uted through a body, it would be inanimate and clearly not an appropriate can- 
didate for a deity. Should the first cause be the other kind of power, the kind that 
is attached to but not distributed through a body — in other words, a soul — it 
would still not deserve the title of deity in the view of the medieval philosophers. 
In their view, a proof of the existence of God has to establish the existence of a 
first cause that is incorporeal and that transcends the universe; an immanent soul 
of the universe would not be the deity. 

The possibility that the mover of the spheres is a power distributed through a 
body had been ruled out by Aristotle and Maimonides through an argument that 
ran: No body can be of infinite magnitude [principle (11)]. Hence, no power 
contained within a body can be infinite [principle (12)]. Finite power is, however, 
insufficient to sustain eternal motion. And the universe and its motion are eternal, 
or at least may be assumed to be so for the sake of the argument [principle (13)]. 
The comprehensive conclusion originally drawn herefrom by Aristotle was that 
the cause of the motion of the spheres and of the universe as a whole must be 
"indivisible, without parts, and without magnitude." 82 The conclusion drawn in 

R0 Crescas explains that principle (9) depends on principle (11). the impossibility of a single 
infinite magnitude. And he refutes principle (1 1); cf. below, pp. 254 ff. Alternatively, he explains, 
principle (9) might be defended through the proposition that an infinite number is impossible. And 
he refutes that proposition too; cf. above, p. 121. 

"Above, p. 249; below, p. 365. 

"Above, n. 49. 



( 

252 Proof from Motion 

/ Maimonides' reformulation of the proof from motion is, more specifically, that 

the first cause of motion cannot be a power contained within, and distributed 
1 through, the body of the spheres. 83 

( Crescas subjects the argument to a most minute critique. He contends that 

three of the four premises in the argument are unfounded. Those premises are 
' the assumption of eternity, the principle that no body can be of infinite magnitude, 

and the principle that a body of finite magnitude cannot contain infinite power. 

Then he contends that even if all the premises are granted, the argument still fails 
' since it confuses two distinct senses of finiteness of power. And finally he contends 

f that the argument can in any event be circumvented by construing the motion of 

the spheres as due not to a power distributed through the body of the spheres or 
( the body of the universe, but rather as due to the nature of the spheres. Crescas 

, thus finds the argument wanting on five separate scores. Keeping the particulars 

of Crescas' discussion in place demands something of the juggler's skill. He is 



( < refuting an argument designed not to prove anything directly, but rather to clim- 

I inate one of three conceivable alternatives to an incorporeal first mover. And his 

attention is focused mainly on the presuppositions underlying the argument's 
( premises. He is, that is to say, examining the presuppositions of the premises of 

I the argument whereby a single alternative to an incorporeal first mover had been 

ruled out. While Crescas' discussion is annoyingly labyrinthine, the circumstance 
f that he focuses on ultimate presuppositions enhances the significance of his cri- 

^ tique. For he successfully exposes both the extent to which the proof of incor- 

poreality, and hence the entire proof from motion, is bound to the presuppositions 
( of Aristotelian physics, and the extent to which those presuppositions can be 

, subverted from within the Aristotelian framework. 



Crescas' critique of principle (13), the assumption of eternity. The eternity 
of the world is, Crescas insists, an indispensable premise in the argument showing 
that the mover of the spheres cannot be a power distributed througli a body. 
Infinite power has to be posited solely in order to explain motion that is sustained 
over an infinite time. Should, on the contrary, the spheres not exist and move for 
an infinite time, a "power which is distributed through a body and therefore 
finite" would suffice to sustain their motion for as long as they do exist. But the 
doctrine of eternity is "undoubtedly false." Consequently, Crescas concludes, the 



( argument showing that the first mover cannot be a power distributed through a 

I body loses an indispensable premise and collapses. 84 

Crescas' thinking on the issue of the eternity of the world was touched upon 
( in an earlier chapter 85 and need not be taken up again here for the following 

( 

( "Above, pp. 245-246. 



M Or lia-Shem. I, ii, 14 (end); 15. 

85 Above, p. 67. Crescas discusses the issue of creation and eternity in Or ha-Shem, III, i, and 
he falls considerably short of showing the doctrine of eternity to be "undoubtedly false." The doctrine 
of eternity would be shown to be false only if the arguments in favor of eternity arc answered, and 



Proof from Motion 



253 



reason: His assertion that a refutation of the doctrine of eternity constitutes a 
refutation of the proof from motion is, although correct from one viewpoint, 
incorrect from another. A successful refutation of the eternity of the world will 
indeed constitute a refutation of the proof from motion when the proof from 
motion is advanced by itself as a categorical demonstration of the existence of 
God; without the premise of eternity, the proof from motion cannot indeed reach 
an incorporeal first mover. But a refutation of the doctrine of eternity does not 
affect the approach of Ibn Tufayl, Maimonides, and Aquinas. Those philosophers 
did not intend the proof from motion to stand by itself, but instead proved the 
existence of God on parallel, alternative tracks, on both the hypothesis that the 
world is eternal and the hypothesis that it is not. The proof from motion has for 
Maimonides and the others no function beyond establishing a first incorporeal 
cause hypothetically — on the first of the two alternatives, that is, the assumption 
that the world is eternal; on the opposite assumption, a first incorporeal cause 
would be inferred directly from the world's having come into existence. 86 Cres- 
cas' refutation of the doctrine of eternity reveals then, if successful, merely that 
the proof from motion cannot single-handedly reach an incorporeal first mover. 
His refutation of the doctrine of eternity does not affect the approach of those 
for whom the proof from motion is intended to do just part of the job, to establish 
the existence of God hypothetically, when the eternity of the world is assumed. 

Crescas' critique of principle (11), the impossibility of an infinite body. The 
impossibility of an infinite body is undoubtedly an indispensable premise in the 
argument showing that the mover of the spheres cannot be a power distributed 
through a body; for if the body of the universe were infinite it would contain 
infinite power, 87 and such power should suffice to sustain the motion of the 
spheres and of the universe as a whole for all eternity. 88 Crescas rejects the 
premise, but not by offering positive arguments to the contrary; he docs not, in 
other words, undertake to demonstrate that the physical universe is in fact of 
infinite magnitude. He only refutes the various considerations whereby the 
"impossibility of an infinite corporeal magnitude" had supposedly been estab- 
lished. 89 Some of those considerations are fully spelled out in Aristotle, while 



counterarguments are adduced to establish creation. As it turns out, Crescas can discover no coun- 
terarguments sufficient, in his view, to establish creation. Of the two sorts of arguments for eternity 
(cf. above, p. 10), Crescas answers those from the nature of the world, but is hard put to answer 
those from the nature of the first cause; cf. Or ha-Shem, III, i, 4; 5. 
86 Cf. above, n. 57 . 

87 Aristotle, Physics VIII, 10, 266b, 6-24, proves that an infinite body must contain infinite 
power. 

"'Or ha-Shem, I, ii, 15. 

m Or ha-Shem, I, ii. 1 (2-4). In Or ha-Shem, I, ii, 1 (1) (= Wolfson, Crescas, pp. 178-191), 
Crescas refutes, as well, the considerations whereby the "impossibility of an infinite incorporeal 
magnitude" has supposedly been established. The issue of an infinite incorporeal magnitude is not, 
however, relevant to the proof from motion. 



254 



Proof from Motion 



others were developed by the Aristotelian commentators, yet without exception 
they are anchored in the most basic presuppositions of Aristotelian philosophy. 90 
Almost all the counter-considerations in Crescas' critique are adaptations of sug- 
gestions that had been made and rejected either by Aristotle himself or by the 
Aristotelian commentators. 

Among the grounds adduced by Aristotle for the impossibility of an infinite 
body had been the definition of body as that which "is encompassed by one or 
more surfaces." Since nothing encompassed can be infinite, the definition, so 
Aristotle's reasoning had gone, entails that every body is finite. 91 Here Crescas 
makes the simple and obvious response that anyone who accepts an infinite body 
will reject the Aristotelian definition. To derive the finitencss of body from the 
definition is to beg the question. 92 

A second ground adduced by Aristotle for the impossibility of an infinite body 
was based on his theory of physical elements. An infinite body, Aristotle had 
reasoned, would have to contain either an infinite number of elements or else at 
least a single element of infinite magnitude. But an infinite body could not, on 
the one hand, consist of an infinite number of elements; for the natures of the 
elements are, he postulated, a subject of human knowledge whereas knowledge 
cannot comprehend an infinite number of items. 93 Nor, on the other hand, could 
any single element be of infinite magnitude. For the infinite element would over- 
power all the other, finite elements and destroy their qualities; it would thereby 
assimilate the other elements to its own nature and give rise to an undifferentiated 
universe. Moreover, an infinite element would occupy all space in the universe, 
leaving no room for other elements. 94 

Crescas' response runs thus: Aristotle's reason for not countenancing an infinite 
number of elements — the infinite is unknowable and hence an infinite number of 
elements could not be a subject of human knowledge — is not cogent, since there 
is no justification for assuming that "the principles qua principles should be 



90 Cf. Wolfson, Crescas, pp. 38-69. 

"Aristotle, Physics III, 5, 204b, 5-6; Or ha-Shem, I, i, 1 (2); Wolfson, Crescas, pp. 150-151 , 
and n. 57. 

92 Or ha-Shem, I, ii, 1 (2); Wolfson, Crescas, 190 -191, and n. 40. In the note, Wolfson quotes 
passages from Narboni and Gersonidcs that may have suggested to Crescas the response he makes; 
and he quotes two surrejoinders to Crescas which were to be advanced by Isaac ben Shem Tob. 

93 0n this point, cf. Aristotle, Physics I, 6, 189a, 12-13; Metaphysics III, 4. 

9i Or ha-Shem, I, i, 1 (2); Wolfson, Crescas, pp. 150-153. Crescas is following Averroes' Middle 
Commentary on the Physics, which combines Aristotle's Physics III, 5, 204b, 10 ff . , and Metaphysics 
XI, 10, 1066b, 26ff.,withf/i.vji'cil,4, 187b, 7-13, and I, 6, 189a, 12-13. The passage in Averroes 
is cited and translated by Wolfson, Crescas, pp. 348-350. 

The possibility of a physical universe constituted of several infinite elements does not seem to 
be covered by the consideration that any infinite element would overpower every finite clement, 
although it is covered by the consideration that an infinite clement would occupy all space in the 
universe. See the passage quoted by Wolfson from Averroes. 



Proof from Motion 



255 



known." 95 Aristotle's reason for not countenancing a single infinite element — 
such an element would overpower the other elements — is also not cogent. For 
the single infinite element might be wholly devoid of positive qualities just as 
the celestial spheres, in the Aristotelian system, are wholly devoid of qualities. 96 
An infinite element need not, moreover, occupy all space; it might extend infi- 
nitely in only one direction leaving room for other elements in the other directions. 

A third ground for the impossibility of an infinite body, formulated not by 
Aristotle but by Averroes, 98 is drawn from the Aristotelian definition of place. 
The place of any object is, according to Aristotle, the inner surface of the body 
or bodies surrounding the object. 99 Clearly, if an object has to be circumscribed 
in order to occupy place, no object occupying place can be infinite. 100 

Crescas responds in two ways. First he notes that even granting the Aristotelian 
definition of place, the definition fails in the case of at least one object in the 
universe: The definition cannot be employed to assign the place of the outer 
celestial sphere or, if we wish, of the universe as a whole. For since nothing, in 
the Aristotelian scheme, surrounds the outer celestial sphere, the place of the 
sphere clearly cannot be the inner circumference of a surrounding body. Some 
sort of ad hoc device must be proposed by Aristotelians in order to assign the 
place of the outermost sphere. And whatever that device should be, it may, 
Crescas contends, equally be employed to assign the place of an infinite body. 
For example, if the place of the outermost sphere is understood, exceptionally, 
to be the outer surface of the body surrounded by the sphere, 101 rather than the 
inner surface of the body surrounding the sphere, then the same may be main- 
tained regarding the place of an infinite body. 102 Such is Crescas' first response 
to the contention that the definition of place excludes the possibility of an infinite 
body. 



n Or ha-Shem, 1, ii. 1 (2); Wollson, Crescas. pp. 192-193. In n. 44, Wolfson cites a number of 
surrejoinders to Crescas. 

"''Or ha-Shem, I, ii, 1 (2); Wolfson, Crescas, pp. 192-193 and nn. 45, 46. Crescas' response 
was already anticipated and rejected by Aristotle, Physics III, 5, 204b, 23-35. 

91 Or ha-Shem. I, ii, 1 (2); Wolfson, Crescas. pp. 192-195 and n. 48. In the note, Wolfson shows 
that Crescas' response was answered, in anticipation, by Averroes. Narboni, and Gersonidcs; and he 
shows that it was also later to be answered by Isaac ben Shem Tob. 

98 Sce the passages from Abraham ibn Daud and Averroes which arc cited by Wolfson, Crescas. 
pp. 352-355. Wolfson suggests that the argument is a development of Aristotle, Physics III, 5, 260a, 
2-8, and Metaphysics XI, 10, 1067a, 30-33. 

"Physics \V, 4. 212a, 20-21. 

100 Or ha-Shem. I, i. 1 (2); Wolfson, Crescas, pp. 152-153. 

101 On the different positions taken by Aristotelians regarding the place of the outer sphere or the 
place of the entire universe, sec Wolfson, Crescas, pp. 432-441. 

,m Or ha-Shem, I, ii, 1 (2); Wolfson, Crescas, pp. 194-195. Crescas means that the place of the 
entire universe would be the center of the universe, or else the inner part of the universe, around 
which the universe is positioned. Sec Wolfson, Crescas. p. 195, n. 52. 



256 



Proof from Motion 



Secondly, and more fundamentally, Crescas responds that the Aristotelian def- 
inition of place leads to so many unacceptable anomalies' 03 that a different def- 
inition, a definition already weighed and rejected by Aristotle, 104 should be pre- 
ferred. Place would best be defined, Crescas maintains, as the interval, or the 
empty void, occupied by an object.' 05 And once this preferable definition is 
adopted, assigning the place of an infinite body is no more difficult than assigning 
the place of a finite body. 106 

A fourth ground for the impossibility of an infinite body, which had been 
adduced by Aristotle, is based not on the general definition of place but rather 
on the Aristotelian theory of natural place, the theory that each of the sublunar 
elements has a proper place to which it naturally travels and which it naturally 
occupies. Aristotle had reasoned that every body composed of elements is either 
heavy, which is equivalent to saying that its natural place is the lower region to 
the exclusion of the upper region; or else it is light, which is equivalent to saying 
that its natural place is the upper region to the exclusion of the lower. But an 
infinite body would be everywhere. It could be neither in the lower region as 
distinct from the upper, nor in the upper as distinct from the lower, and hence 
could be neither heavy nor light. That a body should be neither heavy nor light 
is, however, impossible. And therefore, Aristotle had concluded, an infinite body 
is impossible. 107 Crescas responds, simply, that an infinite body would in fact 
be neither heavy nor light. Such a proposition can hardly be objectionable, seeing 
that the Aristotelian scheme of the universe does recognize bodies that are neither 
heavy nor light, to wit, the celestial spheres. 108 

A fifth ground for the impossibility of an infinite body, which had been offered 
by Aristotle and elaborated by Gersonides, is based again on the Aristotelian 
theory of the natural places of the elements. The reasoning had been: Natural 
motion consists in the elements' returning to their proper places when removed 
therefrom. Now an infinite body would have to contain cither a minimum of one 
element of infinite magnitude or else an infinite number of elements. But, on the 
one hand, a single infinite element cannot be supposed; for an infinite element 
would occupy all space, would always be present in its proper place, and would 
therefore never have the opportunity of undergoing natural motion. On the other 

m 0r ha-Shem, I, ii, 1 (2); Wolfson, Crescas, pp. 194-199. 
104 Cf. Aristotle, Physics IV, 4, 211b, 14-29. 

m> Or ha-Shem, I, ii, 1 (2); Wolfson, Crescas, pp. 194-195, 198-199, and n. 55. Cf. John 
Philoponus, Commentary on Physics, ed. H. Vitclli, Commentaria in Aristolelem Graeca, Vol. XVII 
(Berlin, 1888), pp. 567-569; German translation: W. Boehm, Johannes Philoponus, Ausgewaehlte 
Schriften (Munich, 1967), p. 92. 

106 Or ha-Shem, I, ii, 1 (2). 

""Aristotle, Physics III, 5, 205b, 24-31; Or ha-Shem, I, i, 1 (2); Wolfson, Crescas, pp. 152- 
153, and nn. 65, 67. 

mg Or ha-Shem, I, ii, 1 (2); Wolfson, Crescas, pp. 194-195 and n. 49, where Wolfson refers to 
a similar position taken by Bruno. Aristotle had established that the heavens are neither heavy nor 
light in De Caelo I, 3. 



Proof from Motion 



257 



hand, an infinite number of elements also cannot be supposed. For the infinite 
number of elements would require an infinite number of natural proper places to 
serve as terminuses for their natural motions, whereas the number of natural 
places must be finite, since it is determined by two absolute directions, namely, 
absolute up and absolute down. 109 There is thus, so the conclusion had been 
drawn, no way in which an infinite body might undergo natural rectilinear motion. 

Crescas' response is that the infinite body might after all be composed of an 
infinite number of elements, and nevertheless the proper natural place of each of 
the elements could be assigned. An infinite number of natural places might "exist 
one above the other ad infinitum," 1 10 That is to say, just as both the natural proper 
place of air and the natural proper place of fire are, on the Aristotelian scheme 
of the universe, in the upper region, one above the other, so an infinite number 
of proper places can be conceived in the upper region, one above the other, ad 
infinitum. 

The foregoing ground for the impossibility of an infinite body was the conten- 
tion that such a body could not undergo rectilinear natural motion. That is now 
complemented by a sixth ground, the contention that such a body could not 
undergo the other form of natural motion recognized by Aristotelian physics, 
namely, circular motion. A number of separate considerations came into play 
here. Circular motion, Aristotle had reasoned, takes place around a fixed center; 
but an infinite body would, having no extremities, have no precise center, and 
hence could not undergo circular motion. 1 " Further, circular motion implies a 
spherical body; but an infinite body could not be circumscribed, and hence would 
not have a spherical shape or any other shape." 2 Again, as an infinite body 
revolved, a point infinitely distant from the center would travel over an infinite 
path; but an infinite path is not traversable, and certainly not traversable in a 
finite time." 3 And, yet again, to assume the circular motion of an infinite body 
would involve a paradox. Should an infinite radius be drawn from the center and 
an infinite stationary line drawn parallel thereto, then as the sphere revolved, no 
point could be designated as the first meeting of the revolving radius with the 
stationary line. For no matter how distant a point wc should propose for the first 

'""Aristotle. Physics III, 5, 205a. 10-32; 205b, 31-35; Or ha-Shem, I, i. 1 (3); Wolfson. Crescas, 
pp. 156-159. In nn. 96 and 97, Wolfson quotes the formulations of Averroes and Gersonides. 

""Or ha-Shem, I, ii, 1 (3); Wolfson, Crescas, pp. 202-203. Crescas' response was suggested, 
and rejected, by Gersonides. See Wolfson. ibid., nn. 97, 98. 

'"/Jf Caelo I, 7. 275b, 12-15; Or ha-Shem, I, i. I (4); Wolfson, Crescas, pp. 174-177. The 
point that the inlinitc has no extremities was added by Averroes; cf . the passage referred to by Wolfson 
in n. 158. 

" 2 Aristotlc, De Caelo I, 5, 272b. 17-24; Or ha-Shem, I. i, I (3); Wolfson, Crescas. pp. 172- 
173 and nn. 145 and 148. where Wolfson cites pertinent passages in Averroes. 

'"Aristotle, De Caelo I, 5. 271b, 26— 273a. 6; Or ha-Shem. 1, i, 1 (3); Wolfson. Crescas, pp. 
169-175 and nn. 132, 140, for diagrams. In Aristotle and in Crescas the argument appears several 
times, in several different formulations. 



( 



258 Proof from Motion 

meeting of the two lines, we could always suppose a more distant point where 
the two lines might previously have met. If no point can be designated for the 
first meeting of the radius with the line parallel to it, the radius would be unable 
to revolve." 4 

Crescas responds to all these considerations. An infinite body, he maintains, 
could revolve without a precise center." 5 Further, a revolving body need not be 
spherical or have any other determinate shape. Therefore, although an infinite 
body would lack determinate shape, it would not be prevented thereby from 
undergoing circular motion. 116 Again, no given point in the infinite body would 
traverse an infinite path. For in an infinite body, the "distance [from the center" 7 



/ to points ever more distant on the radius] would increase [infinitely, onlyj in the 

C manner in which number increases [infinitely]," 8 that is to say, without ever 

ceasing to be limited." In other words, although the radius would extend out- 
( wardly without limit, every given point on the radius would remain finitely distant 

^ from the center and hence would describe a finite arc as it revolved. Crescas 

acknowledges that this explanation "is remote from the imagination; nevertheless, 
I reason requires us to accept it." 1 19 As for the supposed paradox in circular motion, 

|. it, according to Crescas, is removed by an Aristotelian observation regarding 

motion and other types of change. Aristotle had noted that there is strictly "no 
I beginning of change," no "first when in which a change occurs," 120 and "no first 

I where to which a change [of place] occurs." 121 Since the beginning of motion 

can in no instance be demarcated, we are, Crescas maintains, not justified in 
f inquiring about the first point where an infinite revolving radius would meet an 

^ infinite line parallel to it. 122 

Finally, a seventh ground for the impossibility of an infinite body, a ground 
i that had been adduced by Aristotle, is based upon the general Aristotelian theory 

. of motion. An infinite body, the reasoning here ran, could not move itself. For 

only animate beings move themselves; animate beings have objects of sense 
( perception outside them and surrounding them; and what is surrounded is finite. 123 



( 

""Aristotle, De Caelo I, 5, 272a, 7-20; Or ha-Shem, I, i, 1 (3); Wolfson, Crescas, 172-173, 
* and p. 387 for a diagram. 

'"Or ha-Shem, I. ii, 1 (4); Wolfson, Crescas, pp. 214-215 and n. 125, where Wolfson refers to 
| a similar position taken by Bruno. 

|; '"'Or ha-Shem, I, ii, 1 (3); Wolfson, Crescas, pp. 212-213. In n. 123, Wolfson cites a passage 

j in Averroes which anticipates Crescas' response, and in n. 122, he refers to a passage in Bruno where 

I that philosopher takes a position similar to Crescas'. 

1 17 As just seen, an infinite body would not in fact have a precise center. 
{ " 8 Cf. Aristotle, Physics 111, 6, 206b, 16-19: The infinite "by way of addition" consists in the 

| possibility of "always taking something ab extra, without, however, exceeding every [determinate] 

i' magnitude." 

I " 9 0r ha-Shem, I, ii, 1 (3); Wolfson, Crescas, pp. 206-213. 

; l2Q Physics VI, 5, 236a, 14-15. m lbid., 236b, 15-16. 

V l22 0r ha-Shem, I, ii, 1 (3); Wolfson, Crescas, pp. 210-213 and n. 120. 

i u3 De Caelo I, 7, 275b, 25-27; Or ha-Shem, I, i, 1 (4); Wolfson, Crescas, pp. 176-177. The 



Proof from Motion 



259 



Nor could an infinite body be moved by something else. For only a commensu- 
rately infinite body would suffice to move the infinite body or act upon it in any 
way whatsoever. And to assume one infinite body's being acted upon by another 
is unacceptable for two reasons. In the first place, the other body would, being 
infinite, possess infinite power; and since the time required to complete an action 
is always inversely proportional to the power performing the action, an infinite 
power would produce instantaneous motion, whereas instantaneous motion is an 
absurdity. 124 In the second place, two actually infinite bodies could not exist 
together; for, as Averroes completed the train of reasoning, an actual infinite 
cannot be added to, and hence no actual magnitude at all can coexist with an 
infinite and add to it. 125 

Crescas responds that an infinite body might in fact move itself without "having 
objects of sense perception outside, surrounding it." 126 His meaning, apparently 
is not that the infinite body would be animate, but that it would be moved by a 
physical power contained within it 127 or else by its own nature. 128 As for the two 
reasons why an infinite body could not be moved by another infinite body, Crescas 
responds, in the first place, by proposing a hypothesis going back to John Phil- 
oponus. The hypothesis will be examined more fully in later connection, 129 but 
stated briefly it is this: Although the time required for any given operation is 
undoubtedly reduced as the power performing the operation is increased, every 
operation has, intrinsic to itself, a certain minimum "basic time," which is irre- 
ducible no matter how great a power might be brought to bear. An infinite power 
would accordingly perform each operation not instantaneously, but in the irreduc- 
ible time intrinsic to the operation. Thus, to assume an infinite body containing 
infinite power would not entail the assumption of instantaneous action. 130 In the 
second place, Crescas responds that two infinite bodies could after all exist side 
by side. For infinite bodies need not be infinite in every direction; and the infinite 
may be added to on its finite side. 131 

The upshot of Crescas' critique of the various grounds whereby Aristotle and 
his followers had ruled out an infinite body is that an infinite body might in fact 



consideration that "animate bodies have objects of sense perception outside them and surrounding 
them, etc." is Averroes' addition. Sec the passage referred to by Wolfson, ibid., n. 160. 

for Maimonides' argument against an animate body's moving itself for all eternity, cf. above, 
pp. 246-247. 

'"Aristotle, De Caelo I, 7, 274b, 33— 275b, 4; Or ha-Shem, I, i, 1 (3); Wolfson, Crescas, pp. 
164-169. 

'"Aristotle, De Caelo I, 7, 275b, 27-29; Or ha-Shem, 1, i, 1 (4); Wolfson, Crescas, pp. 176- 
177 and n. 160, where Wolfson refers to the pertinent passage in Averroes. 
'^Or ha-Shem, 1, ii, 1 (4); Wolfson, Crescas, pp. 214-215. 
l27 Cf. below, pp. 264 - 265. 

,28 Cf. below, pp. 269-270. '"Below, pp. 262-263. 

,30 Or ha-Shem, I, ii, 1 (3); Wolfson, Crescas, pp. 204-205. 

"'Or ha-Shem, I, ii, 1 (4); Wolfson, Crescas, pp. 214-215; above, n. 97. 



260 



Proof from '.lotion 



exist. Such a body could either be comprised of natural elements of the sublunar 
kind or else have the nature of the celestial region, the latter possibility, the 
existence of an infinite celestial region, clearly being intended more seriously. 132 
An infinite celestial body would not, Crescas has explained, be strictly spherical 
nor would it have any other definite shape. It would rotate without having a 
precise fixed center. It would be free of the qualities of the sublunar elements 
and would possess neither weight nor lightness. Its place could be assigned in 
whatever way Aristotelians might choose to assign the place of the outer celestial 
sphere which, in the Aristotelian system, is finite. And an infinite celestial body 
could contain infinite power distributed throughout it. It could, consequently, 
sustain its own circular motion for all eternity. 

The object of Crescas' critique in the present section was, as must be kept in 
mind, to refute just a single premise in the argument intended to show that the 
motion of the spheres cannot be due to a power distributed through the body of 
the spheres. The argument had been that since all bodies are finite, every power 
distributed through a body must be finite; and therefore the eternal motion of the 
universe cannot be sustained by a power distributed through a body. Crescas, in 
the stage of his critique we have been examining, has rejected the premise that 
an infinite body is impossible. If an infinite body is after all possible, an infinite 
power distributed through a body is also possible. Even granting then that the 
motion of the universe is eternal and must be sustained by infinite power, the 
eternal motion of the universe might be sustained through the physical power 
distributed through the infinite body of the spheres or the infinite body of the 
universe. 

Crescas' critique of the principle that a finite Lody can contain only finite 
power. Crescas has, he is confident, refuted two premises in the argument intended 
to show that the motion of the spheres cannot be due to a power distributed 
through a body; those premises are the doctrine of eternity and the impossibility 
of an infinite body. A third premise in the argument is the principle that a finite 
body can contain only finite power. The premise is again indispensable. For — 
conceding now the finiteness of the body of the universe, as well as the eternity 
of the world — if a finite body could contain infinite power, such a body could 
contain within itself the power needed to sustain motion for an infinite time. The 
first cause of motion in the universe, even if motion were eternal, could be an 
infinite physical power distributed through the finite body of the universe. 

As is his custom, Crescas undertakes not to disprove the principle at issue, but 
only to refute the philosophic grounds for it. He does not, in other words, propose 

132 An infinite body comprised of elements of the sublunar kind could be assumed only if the 
distinction between the sublunar and celestial regions were eliminated; and Crescas docs not seem 
willing to countenance that step seriously. The picture he is suggesting is that of a celestial region, 
dotted with stars and spheres, which encompasses the sublunar world and extends outwardly without 
limit. 



Proof from Motion 



261 



to show that the principle is false and that finite bodies do in reality contain 
infinite power. He proposes to show merely that the principle is unfounded and 
that finite bodies might contain infinite power. 

Aristotle's reasoning in support of the principle had been more intricate than 
the reasoning Maimonides was subsequently to employ, and it is with Aristotle's 
reasoning that Crescas is concerned. Aristotle had undertaken to reduce the 
assumption of an infinite power contained within a finite body to absurdity. On 
the one hand, he had held, an infinite power contained within a finite body could 
not perform its operation instantaneously, since an instantaneous physical oper- 
ation is absolutely inconceivable. And, on the other hand, he had contended, an 
infinite power contained within a finite body could not operate over a finite time 
span either; for there is no time span that might be assigned for its operation. 
The inconceivability of instantaneous physical operations was so obvious to Aris- 
totle that he could take it for granted. 133 His efforts were directed towards dem- 
onstrating that there is no time span assignable for the operation of an infinite 
power contained within a finite body. 

Aristotle's reasoning ran: "Suppose z to be the time required for the infinite 
power to perform [a certain operation]." Since a finite power would surely require 
more time, we may suppose "z plus y to be the time required for a given finite 
power to perform the same operation." Now the time required to complete a fixed 
amount of work is inversely proportional to the power applied. 134 Should the 
given finite power be increased, the time — z plus y — needed to complete the 
operation in question would decrease. Eventually, by continuing to "increase the 
finite power ... we can reach the level where it performs the operation in time 
z." 135 For example, if z'plus y were twice z, by doubling the finite power we 
would reduce the time required to perform the operation by half. "The finite 
power would then require the same time that the infinite power was supposed to 
require." That a finite power and an infinite power should perform the same work 
in the same time — that they should, in other words, operate at the same velocity — 
is, however, absurd. And yet, no matter how brief a time span and how rapid a 
velocity might be supposed for the operation of the infinite power contained 
within the finite body, no matter how brief z might be, the situation would be 
unchanged. A finite power would in each instance be discoverable which can 
perform the same operation in the same time, at the same velocity; and a situation 
of the sort is absurd. To suppose that the infinite power performs the operation 
in no time, at an infinite velocity, was excluded at the outset as absolutely incon- 
ceivable. The conclusion accordingly drawn by Aristotle is that no time span or 
velocity whatsoever can be assigned for the operation of the infinite power. If 
there is no time span or velocity in which an infinite power contained within a 

l33 Cf., however, De Sensu 6, 447a, 1-3. 

,34 Cf. Physics VU, 5. z+ , 
l35 The level will be reached when the finite power is multiplied by . 



262 



Proof from Motion 



finite magnitude can act, the supposed power simply could not act. And a power 
that docs nothing is not a power. Thus a finite body can nowise contain an infinite 
power. 136 

Such were Aristotle's grounds for the principle that a finite body can contain 
only finite power. The key to Aristotle's reasoning is the rule that the time required 
for any operation is inversely proportional t^ the power applied, and Crcscas 
addresses himself to that rule. Crescas does not dispute the empirical fact that as 
the power performing a given operation is increased, the time required to com- 
plete the operation decreases. He maintains, though, that the empirical fact docs 
not necessarily imply that power and time stand exactly in inverse proportion. 
At least as plausible is a different hypothesis, going back to John Philoponus, 137 
which was mentioned earlier in another connection. 138 

The alternate hypothesis is that each given physical operation has, intrinsic to 
itself, a certafii minimum basic time span (zeman shorshi) for its occurrence, a 
time span that is irreducible no matter how great a power may be brought to bear. 
Stated otherwise, every physical operation has its own intrinsic maximum veloc- 
ity which cannot be exceeded. When an operation is performed by a finite power, 
the minimum time span does not suffice, and additional time is required, with 
the amount of the addition dependent on the power. What is inversely proportional 
to the power performing a given operation is not the total time required for the 
operation, but rathe the increment over the minimum time. For example, should 
the power be doubled, what would be reduced by half is not the total time of the 
operation, but the increment. As long as the power is multiplied finitely, the 
increment will be divided finitely, with some of it always remaining; and when 
the power reaches infinity the increment will be reduced to zero. "The infinite 

l36 Physics VIII, 10, 266a, 24 — 266b, 6. There is an embarrassing side to Aristotle's argument. 
The argument appears to pro.e not merely that no infinite power in a finite magnitude could perform 
any operation, but more generally that no infinite power whatsoever could, and hence that no infinite 
power whatsoever can exist. Such, of course, is not the conclusion Aristotle was aiming at. He wished 
to affirm, on the contrary, that an infinite power does exist, to wit, the infinite power maintaining the 
heavens in motion, and he traced that infinite power to an incorporeal being. To salvage Aristotle's 
proof of the principle that a finite magnitude cannot contain infinite power, some sort of distinction 
would have to be drawn. The inconccivabilty of finite and infinite powers' performing the same 
amount of work in the same time would have to be restricted in some way to powers in bodies, and 
more precisely, as Aristotle formulates his proof of the principle, to powers in finite bodies. The 
power of an incorporeal being could then be subject to different rules, so that it could, with no 
absurdity ensuing, be assumed infinite and yet able to mete out its activity at the same velocity as a 
finite power contained in a body. A distinction of the requisite sort, between powers in bodies and 
powers of incorporeal beings, seems to be adumbrated by Simplicius, Commentary on Physics, pp. 
1321, 1339; and such a distinction is worked out by Averroes, Long Commentary on the Physics, 
VIII, comm. 79. 

'"For Philoponus, see M. Clagett, The Science of Mechanics in the Middle Ages (Madison, 1959), 
pp. 433-434. The hypothesis was advanced in the Middle Ages by Ibn Bajja and was known to 
Crcscas through Averroes' critique of that philosopher; See Wolfson, Crescas, pp. 403-408. 

138 Above, p. 259. 



Proof from Motion 



263 



| power] would thus require no time beyond the irreducible minimum in order to 
produce motion, whereas every finile | power | will require ..onie increment over 
the minimum." ly ' 

The hypothesis outlined avoids the twin horns of the dilemma posed by Aris- 
totle. Instantaneous physical operations and infinite velocity, regarded by all as 
absolutely inconceivable, are avoided, since a finite time span and a finite velocity 
are assigned foi every operation performed by the infinite power contained within 
a finite body. And the absurdity of a finite power's operating at the same velocity 
as an infinite power is also avoided. The infinite power and it alone would com- 
plete each operation within the irreducible minimum time intrinsic to the oper- 
ation, whereas every finite power would require an increment over the mini- 
mum. 140 Aristotle's proof of the principle that a finite body can contain only finite 
power consisted in showing that no time span could be assigned for the operation 
of an infinite power contained within a finite body. By explaining how a time 
span can be assigned, with no absurdity ensuing, Crescas has refuted Aristotle's 
proof of the principle. 

Crcscas' object, we must recall once again, is to invalidate the argument show- 
ing that celestial motion and, in general, motion in the universe cannot be due 
to a power distributed through a body. In the present stage of his critique, Crescas 
has refuted the Aristotelian grounds for one more premise in the argument, the 
principle that a finite body can contain only finite power. The effect of his refu- 
tation is to permit ascribing infinite power to the finite universe, and hence to 
permit ascribing to the finite universe sufficient power for sustaining its own 
motion over an infinite time. 141 

The distinction between finiteness in respect to intensity andfinileness in respect 
to continuity. Crescas has refuted the grounds for three of the four premises in 
the argument showing that the ultimate mover of the spheres cannot be a power 

" 9 0r ha-Shem. I, ii, 8; Wolfson, Crescas. pp. 270-271. 

l40 Crcscas, Or ha-Shem, 1, ii, 8, also proposes a variation of the hypothesis just outlined. A finite 
power, he writes, may in some instances be conceded to require no increment above the minimum 
time. In other words, a finite power may be conceded to move some objects in the same minimum 
lime and at the same velocity that an infinite power requires. Yet the finite power will not necessarily 
be as efficacious. For even acknowledging those instances, we may conceive of a larger object to be 
moved, an object of such magnitude that the finite power would no 1 gcr move it in the minimum 
time but would now require an increment. The infinite power would, by contrast, never require more 
than the minimum time to move its objects, no matter how great they might be. According to Crcscas. 
this variation of the hypothesis again invalidates Aristotle's proof of the principle that a finite body 
cannot contain infinite power. Aristotle's contention, as Crcscas understands it, turns on the absurdity 
of a finite power's being fully as efficacious as an infinite power. In the present variation of the 
hypothesis, a velocity can again be assigned for the operation of the infinite power without that 
absurdity's ensuing, for although the finite power would in some instances operate as fast as the 
infinite power, it would not do so in all instances and hence would be less efficacious. 

[4 '0r ha-Shem, 1, ii, 15. 



264 



Proof from Motion 



distributed through the body of the spheres. He raises an additional objection: 
Even conceding the premises — the eternity of the universe, the finiteness of the 
body of the universe, and the principle that a finite body can contain only finite 
power — the argument is still invalid. For the argument ignores the distinction 
between finiteness in respect to intensity and finiteness in respect to continuity. 
The distinction between these two senses of finiteness was suggested to Crescas 
by Averroes, but Crescas utilizes the distinction differently. 142 

"The term infinite" he explains, "clearly may be used in two senses, both with 
respect to intensity and also with respect to time [or continuity]." The principle 
that a finite body can contain only finite power, granting now that Aristotle had 
successfully demonstrated the principle, affirms no more than that a finite body 
cannot contain power "infinite in respect to intensity." A power finite in intensity 
might, however, be infinite in respect to continuity, and so might suffice to sustain 
the motion of the heavens eternally. 143 The principle that a finite body contains 
only finite power therefore does not, Crescas objects, lead to the conclusion that 
the power distributed through a finite body would be insufficient to sustain eternal 
motion. 144 

Crescas presumably means the following: Aristotle's reasoning in support of 
the principle that a finite body can contain only finite power consisted in showing 
that no time span, in other words, no velocity, can be assigned for the operation 
of an infinite power contained in a finite body. Whatever time span might be 
proposed, a finite power could be discovered which would perform the same 
operation in the same time and at the same velocity; 145 and that would be absurd. 
But in Aristotelian physics, the intensity, not the continuity, of a power determines 
velocity. 146 That is to say, increasing the intensity of a power increases the velocity 
of its operations and decreases the time span required to complete each operation; 
increasing the time over which the power is applied has, by contrast, no effect 
on the velocity. To assert that increasing the time over which a power operates 
decreases the time required to complete the operation would, in the Aristotelian 
scheme, be nonsensical. 147 Aristotle's reasoning in support of the principle that 
a finite body can contain only finite power — reasoning which consisted in show- 
ing that no time span can be assigned for the operation of an infinite power 
contained in a finite body — is comprehensible, then, solely in reference to finite- 
ness of intensity. Aristotle can at best be understood to have shown solely that 



142 Cf. below, p. 323. The distinction between infinite in intensity and infinite in continuity had 
also been drawn by Avicenna, Najat, p. 128. 

M Or ha-Shem, I, ii, 8; Wolfson, Crescas, pp. 272-273. 

,44 Or ha-Shem, I, ii, 15. Cf. the account of Buridan's discussion of the principle in A. Maier, 
Melaphysische Hintergruende der Spaetscholastischen Naturphilosophie (Rome, 1955), pp. 257 ff. 
l45 Cf. above, pp. 261-262. 

'"Continuing to apply power merely sustains velocity and does not produce acceleration. Cf. 
above, nn. 16, 29. 
l47 See previous note. 



Proof from Motion 



265 



no conceivable time span and no velocity can be assigned for the operation of a 
power of infinite intensity contained within a finite body, since a power of finite 
intensity could always be discovered which would perform the same operation 
in the same time and at the same velocity. Aristotle's reasoning nowise explains 
why a velocity cannot be assigned for the operation of a power of infinite con- 
tinuity contained within a finite body. Aristotle may, of course, have regarded 
intensity and continuity as convertible, so that a power infinite in one respect 
would be infinite in the other as well. But he did not affirm, let alone defend, 
the convertibility of intensity and continuity. Hence Crescas' objection: Even 
granting the principle that a finite body contains only finite power, the grounds 
whereby Aristotle supported the principle can establish no more than that a finite 
body contains power finite in intensity; and a power finite in intensity might very 
well be infinite in continuity. The body of the spheres and the body of the universe 
might thus be finite in magnitude; the power contained within the body of the 
spheres and the body of the universe might be finite in intensity; and yet the 
power of finite intensity contained within the universe might suffice to sustain 
the motion of the spheres for all eternity. The ultimate cause of eternal motion 
may be construed as a physical power contained within and distributed through 
the finite body of the universe, even if all the premises in the Aristotelian argu- 
ment intended to exclude the possibility be granted. 

Circular motion may be natural to the spheres. Crescas raises yet another 
objection to the argument whereby Aristotle tried to show that the mover of the 
spheres cannot be a power distributed through the body of the spheres. The 
argument, he maintains, can be completely circumvented by construing the motion 
of the spheres as "natural to the body of the spheres just as rectilinear motion is 
natural to the elements." 148 Crescas' objection must be read against the back- 
ground of positions taken by Aristotle and by ancient and medieval philosophers 
standing in the Aristotelian tradition. 

"Nature," a well-known Aristotelian formula affirms, is "a certain principle 
and cause of [a thing's] being moved and being at rest." 149 The rectilinear motion 
of the four sublunar elements toward their proper places had, in accordance with 
the formula, been characterized by Aristotle as natural, inasmuch as the elements 
have within themselves "a principle of motion." 150 The circular motion of the 
celestial spheres had similarly been characterized by Aristotle as natural, inas- 
much as the spheres too have a "principle of motion in themselves." 151 Yet side 
by side with the characterization of the motion of the elements as natural, Aristotle 



'""Or ha-Shem, I, ii, 8; Wolfson, Crescas, pp. 272-273. 
'"Physics II, I, 192b, 21-22. 

,50 De Caelo III, 2, 301a, 21 ff.; IV. 3, 310b, 24-25. In the former passage, Aristotle speaks of 
the "natural momentum (po-nri)" of the elements for moving up or down. 
"'Ue Caelo I, 2; II, 2, 285a, 30. 



< 

\ 

( 

266 Proof from Motion 

{ 

f had undertaken to establish that the sublunar elements "are not moved by them- 

selves." 152 That is to say, although the elements are moved owing to their nature, 

i they are not moved by their nature. How the nature of the elements contributes 

to their motion is not clear in Aristotle. 

1 Convinced as he was that the elements do not move themselves, Aristotle was 



( hard put to discover what, external to the elements, does produce their rectilinear 

^ motion. The position he arrived at was that the factor originally bringing a given 

element into existence should be taken as the cause of the motion of the element 
( to its proper place; for example, the motion of a portion of air to the upper region 

( is to be ascribed to the factor bringing the portion of air into existence. And, 

Aristotle further explained, in instances where an element has already been brought 
( into existence but is somehow prevented from moving to its proper place, the 

/ factor removing the obstacle is to be identified as the cause of motion. 153 These 

' explanations account at best for the initiation of the movement of the elements, 

f In the Aristotelian system, motion has to be sustained as long as it continues, 



and Aristotle's explanations fail to account for the continuation of the elements' 
motion as they make their way toward their proper places. 154 

Having traced the natural rectilinear motion of the elements to a cause distinct 
from them, Aristotle undertook to trace the natural circular motion of the celestial 
spheres to a cause distinct from the spheres. But since he understood the motion 
of the spheres to be eternal he did not, in their case, seek a cause that initiates 
motion. Rather, he sought the cause that sustains the motion of the spheres. In 
certain passages Aristotle explained that the motion of the celestial spheres is 



( sustained by a soul. 155 In other passages he explained, as we have seen in the 

present chapter, 156 that celestial motion is sustained by an entity that has no 

f magnitude and is nowise attached to a sphere, in other words, by a purely incor- 

/ poreal being. Although the two accounts may appear divergent to modern schol- 

ars, 157 they were read by the ancient and medieval commentators as 

( complementary. 158 

/ According to Aristotle, then, the elements and spheres arc in motion — the 



former rectilinearly and the latter circulary — owing to their nature; yet the cause 
producing the motion of elements and spheres is not their nature but something 

'"Physics VIII, 4, 255a, 1 ff. 

'"Ibid., 256a, 1-3; De Caelo IV, 3, 311a, 9-11. 

" 4 Cf. G. Secck, "Leicht-schwer und der Unbewcgtc Bewegcr," in Nalurphilosophie beiAhsloleles 
und Theophrast, ed. I. Dueling (Heidelberg, 1969), pp. 214-215. 
'"De Caelo II, 2, 285a, 29-30; 12, 292a, 20-21. 
156 Above, n. 49; also Metaphysics XII, 7. 

157 Cf. H. von Arnim, "Die Entstehung der Golteslehre des Aristolclcs," reprinted in Metaphysik 
und Theologie des Aristoteles, ed. F.-P. Hager (Darmstadt, 1969), pp. 1-74; W. Guthrie, "The 



I Development of Aristotle's Theology," Classical Quarterly, XXVII (1933), 162-171; Ross's intro- 

duction to his edition of the Physics, pp. 94-100. Guthrie finds a less radical development than does 
/ von Arnim, and Ross finds a still less radical development than Guthrie. 

"'Alexander already read Aristotle in that way; cf. Aporiai, ed. I. Bruns, Commemaria in Aris- 
\ toteiem Graeca, Supplementary Vol. II/2 (Berlin, 1892), II, i, p. 3. 



Proof from Motion 



267 



external. In the case of the rectilinear motion of the elements, Aristotle identified 
the cause responsible for the inception of motion, the initiating cause. The motion 
of a given portion of an element is initiated by the factor bringing the portion of 
the element into existence, or else by the factor removing the obstacle to its 
motion. Aristotle's writings reveal no attempt to identify what it is that sustains 
the motion of the elements as they continue to move. In the case of the motion 
of the spheres, by contrast, Aristotle had to identify only the cause responsible 
for the continuation of motion, that is, the sustaining cause. The circular motion 
of each sphere is sustained — as ancient and medieval philosophers read Aristo- 
tle — by an incorporeal being that interacts with the soul of the sphere. 

Despite Aristotle's insistence that the sublunar elements do not move them- 
selves but are moved by something external to them, a number of ancient and 
medieval students of his philosophy did discover a motive factor within the 
elements. And the motive factor was found to be, appropriately enough, the 
nature of the elements; nature had, after all, been characterized by Aristotle as a 
cause and principle of motion, with no exposition of the way in which it is a 
cause and principle. 

The earliest work, as far as I know, that construes the nature of the elements 
as a motive cause is John Philoponus' commentary on Aristotle's Physics. Aris- 
totle's definition of nature — the "principle and cause of [a thing's] being moved 
and being at rest" — was interpreted by Philoponus as meaning that natural sub- 
stances such as the elements have the cause of their motion "in themselves." 
Hence when a heavy or light object is "released," its movement towards its 
proper place is brought about by its own nature and not by anything outside it. 159 
Philoponus probably did not wish, hereby, to reject Aristotle's identification of 
the factor bringing the element into existence as the initiator of the element's 
motion. 160 He wished merely to ascribe to the nature of the element a function 
not treated by Aristotle, namely, the function of sustaining motion once the 
clement exists and is free to move. 

A statement to the effect that the elements are moved by their own nature 
appears, in the Middle Ages, in a corpus of Arabic works attributed to Alexander 
of Aphrodisias; unfortunately, however, the exact intent there remains obscure. 161 
Subsequently, Avicenna, followed by Ghazali in his restatement of Avicenna's 
philosophy, asserted that the "nature" of the sublunar elements is the "cause" of 
their motion. And the context wherein Avicenna made the assertion indicates 162 

'"Philoponus, Commentary on Physics, p. 195; cf. S. Pines, "A Refutation of Galen by Alex- 
ander," lsis, LII (1961), 48. Also cf. Simplicius, Commentary on Physics, pp. 1217, 1220, and Pines, 
ibid., pp. 39-40. 

l60 He does eliminate the factor removing the hindrance to motion as in any sense a motive factor. 

'"'Mabtidi' al-Kull. pp. 253-254. Cf. Pines, "A Refutation of Galen by Alexander." pp. 27, 32, 
42-48; H. Wolfson, "The Problem of the Souls of the Spheres," reprinted in bis Studies in the History 
of Philosophy and Religion, Vol. 1 (Cambridge, Mass., 1973), p. 29. 

'"Since Avicenna and Ghazali mention no cause apart from nature, they apparently recognized 
no such cause. 



268 



Proof from Motion 



that he, followed by Ghazali, was construing the nature of the elements as the 
exclusive cause of their motion. In other words, each element is not merely 
sustained in motion, but is set in motion as well, by its own nature. 163 Avicenna 
acknowledged, of course, that the elements arc brought into existence by some- 
thing different from themselves. 164 Yet he demurred, so it appears, at Aristotle's 
identification of the factor bringing an element into existence as the initiator of 
its motion. 165 

Avicenna's position that the nature of each element is responsible for all aspects 
of the element's motion was not generally adopted. But from the time of Averroes 
onward, the nature of the elements was repeatedly credited with an active function 
in the elements' natural motion. The motion of an element was commonly said 
to be initiated, indeed, by the factor bringing it into existence; but once the 
element does exist, the factor sustaining its motion until it reaches its proper 
place was understood to be the element's inner nature. A theory of the sort was 
offered in divers versions by Averroes, 166 by the Jewish philosopher Isaac 
Albalag, 167 and by a line of Latin writers extending from Albertus. Magnus to 
the fourteenth-century Scholastics. 168 

Crescas for his part follows the lead of Avicenna in dispensing completely 
with external factors to explain the motion of the elements to their proper place. 
What causes each element to move up or to move down to its proper place is, 
Crescas maintains, "the proper form" of fhe element, which operates through the 
"medium of the power . . . called nature." 169 That is to say, the form of the 
clement, operating through the nature of the element, is the sole cause of the 
element's motion. 

Such was Crescas' position on the natural motion of the sublunar elements and 
his precedents. 

There were in addition at least two precedents, both of them conscious depar- 
tures from Aristotle, for explaining the circular motion of the celestial spheres 
as due exclusively to the nature of the spheres, thereby dispensing with the soul 



163 Avicenna, Najat, pp. 108-109; Ghazali, Maqasid, p. 239. Cf. Shem Tob's commentary on 
Guide, II, introduction, prop. 17; Wolfson, Crescas, p. 673; Pines, "A Refutation of Galen by 
Alexander," p. 51. 

164 Thcy are brought into existence through a form emanated from the active intellect. 

165 Therc is a bit of hairsplitting here, since the cause of the existence of the clement and of its 
nature would have to be recognized as an ultimate cause of the motion of the clement. 

l66 Epitome of De Caelo, in Rasd'il Ibn Rushd (Hyderabad, 1947), p. 68; Long Commentary on 
DeCaelo, in Aristotelis Opera cum Averrois Commentaries, Vol. V (Venice, 1562), III, comm. 28. 

'"Isaac Albalag, Sefer Tiqqun ha-De'ol (Commentary on Ghazali's Maqasid), ed. G. Vajda 
(Jerusalem, 1973), pp. 102-103. 

168 A. Maier, An der Grenze von Scholastik und Naturwissenschaft (Rome, 1952), pp. 156-170. 
Maier finds that until Duns Scotus, the nature of the elements was construed as only an "instrument" 
of the factor bringing the element into existence. Also cf. J. Weisheipl, "The Principle Omne quod 
movetur ab alio movelur," Isis, LVI (1965), 39-41. 

I69 0r ha-Shem, I, i, 17; Wolfson, Crescas, pp. 298-299 and n. 7. Aristotle, De Caelo III, 2, 
301b, 18, distinguishes nature from power. 



Proof from Motion 



269 



of the spheres and with the incorporeal intelligences. Aristotle, as already seen, 
had stated that the celestial spheres perform their circular motion thanks to their 
nature. 170 Undoubtedly with that statement in mind, John Philoponus contended 
in a nonphilosophic tract that "God" may very well have furnished the heavenly 
bodies with "a power of movement" analogous to the "momentum" 171 whereby 
the sublunar elements travel to their proper places. 172 In other words, the power 
actively moving the spheres may be their nature. Ghazali offered a similar sug- 
gestion when he turned from his restatement of Avicenna's philosophy to a cri- 
tique of Avicenna. The body of the heavens, Ghazali now contended, may "have 
been created, . . . possessing in itself a factor that necessitates circular motion." 
The factor in question would be "a principle of motion," that is, a nature, 173 and 
its character would be "analogous to what the philosophers believe regarding the 
downward motion of a stone." 174 In these passages, Philoponus and Ghazali 
exploited the Aristotelian notion of nature to develop an anti-Aristotelian stand. 
They assumed the creation of the heavens by God. And they contended that, once 
created, the heavens may well be moved — whether set in motion or only sustained 
in motion is not made explicit — exclusively by their nature. 

In the passages quoted, Philoponus and Ghazali had exchanged philosophic 
for theological garb and were approaching the issue of celestial motion theisti- 
cally. Their contention was that once God created the celestial spheres, the nature 
of the spheres could suffice to move them. Crescas is approaching the issue of 
celestial motion from a different direction, and as a result he goes further. He, 
like Aristotle and Maimonidcs, is addressing the question: What could sustain 
the motion of the celestial spheres on the assumption that the spheres and their 
motion are eternal? The theory Crescas advances is that the eternal motion of 
the spheres can be construed as "natural to the spheres," just as the "rectilinear 
motion of the elements" is "natural" to the elements. 175 That is to say, just as 
the rectilinear motion of the sublunar elements to their proper places can, on the 
position of Crescas' outlined previously, 176 be ascribed exclusively to the nature 
of the elements, with no role assigned to external factors, so too the circular 



""Above, pp. 265-266. 
,7 'Cf. above, n. 150. 

m De Opi/icio Mundi. cd. G. Rcichardt (Leipzig, 1897). p. 29, and also pp. 231-232; German 
translation: Hochm, Johannes I'hiloponus, Ausgewaehlle Schriften. pp. 333-334, 335-336. Cf. 
Wolfson, "The Problem of the Souls of the Spheres," p. 26; Pines, "A Refutation of Galen by 
Alexander," p. 50. 

l73 Cf. above, n. 149. 

174 Ghazali. Tahafut al-Falasifa, ed. M. Bouyges (Beirut, 1927), XIV, §11; English translation 
in Averroes' Tahafut al-Tahafut, trans. S. van den Bergh (London, 1954), p. 290. 

I75 0 ha-Shem, I, ii, 15; also ibid., I, i, 6; Wolfson, Crescas, pp. 236-237. Crescas, ibid., I, ii, 
15, acknowledges that the analogy is not exact since "in the case of the elements, to rest in their 
proper place is natural, whereas in the case of the celestial bodies, to move in their proper place is 
natural." 

,76 Above, p. 268. 



( 

i 

( 

270 Proof from Motion 

; 

, motion of the spheres can be ascribed exclusively to the nature of the spheres, 

and no additional factor need be introduced. Ascribing the motion of the spheres 

( exclusively to their nature would sidestep the argument to the effect that the 

power contained within the finite body of the universe cannot sustain eternal 

1 motion. It may be granted, Crescas is contending, that motion is eternal, that the 

( universe is finite, that the power contained within the body of the universe is also 

/ finite, even that the power contained within the universe is finite in respect to 

continuity. Conceding all this, the argument that a power distributed through a 

( finite body would not suffice to sustain eternal motion can yet be circumvented 

by construing the motion of the spheres as natural to the spheres. The nature of 
the spheres is present and operative as long as the form is present, 177 that is, as 

( long as the spheres exist, Hence the nature of the spheres can move the spheres 

as long as they exist, whether it be for a finite or infinite time. And no incorporeal 

' mover beyond the spheres need be posited. 



Resume. We have been considering the argument intended, in Maimonides' 
/ version of the proof from motion, to eliminate one conceivable explanation of 

the motion of the celestial spheres, the alternative that the cause of celestial 
motion is a power distributed through a body. The argument had been that no 
( body can contain infinite power, whereas finite power cannot sustain eternal 

motion. 

* Crescas undertook to refute three of the premises in the argument. A refutation 

( of the first premise, the doctrine of eternity, would not affect the final outcome. 

For the function of the argument was to serve as a step in an overall proof of the 
' incorporeality of the first cause, and denying eternity and thereby affirming a 

( beginning of the world leads to the incorporeality of the first cause by another 

route. 178 The refutation of a second premise, the principle that only finite bodies 
^ exist, would invalidate the argument. For if the heavens and the universe as a 

( whole should be infinite, they would contain infinite power, and infinite power 

would suffice to sustain eternal motion. The refutation of a third premise, the 
' principle that a finite body contains only finite power, would also invalidate the 

( argument. For if a finite body should contain infinite }owcr, that power, again, 

would suffice to sustain eternal motion. 
( Having objected that the argument is invalid since two — or, including the 

|/ doctrine of eternity, three — premises are unfounded, Crescas contends further 

j that the argument remains invalid even when all the premises are granted. Aris- 

I ' totle's reasoning supporting the principle that a finite body contains only finite 



( power would, if correct, establish, no more than that a finite body contains power 

j finite in intensity. Aristotle's reasoning nowise establishes that a finite body con- 

j ( tains power finite in continuity. Therefore, even conceding the principle that a 

\( 

if l77 Sce above, n. 169. 

I 178 Above, n. 57. 

|f 

\< 



Proof from Motion 



271 



finite body contains only finite power, such power, finite though it is in intensity, 
may still suffice to sustain eternal motion. Finally, Crescas contends, even if yet 
another concession is made and the validity of the entire argument is granted, 
the argument can still be circumvented by identifying the cause of the eternal 
motion of the spheres as their nature. 

(Hi) The alternative that the first mover of the spheres is a power 
attached to, but not distributed through, the body of the spheres 

The first conceivable explanation of the movement of the spheres had been the 
hypothesis that the spheres are moved by another body. The argument whereby 
that alternative was ruled out in Maimonides' version of the proof from motion 
and Crescas' reply thereto were examined earlier. 179 The second conceivable 
explanation was the hypothesis that the mover of the spheres is a power distributed 
through a body. The argument whereby that alternative was ruled out and Crescas' 
intricate reply have just been examined. The third conceivable explanation is the 
hypothesis that the mover of the spheres is a power present within, although not 
distributed through, the body of the spheres, in effect, a soul of the spheres. 

Maimonides' version of the proof from motion had eliminated the third alter- 
native through an argument that ran: The motion of the spheres is by hypothesis 
eternal. But a power in a body, even if, like a soul, not distributed throughout 
the body, would undergo motion accidentally; the accidental has the possibility 
of not existing, and hence is not eternal; the motion of a power attached to a 
body must, therefore, eventually cease; and the motion of the body inextricably 
associated with a power whose motion must cease must also eventually cease. 180 

The argument has a flaw even when considered from a purely Aristotelian point 
of view, inasmuch as it ignores the differing senses of accident. Accident, it is 
true, was defined by Aristotle as "what is present in something ... but neither 
necessarily so nor for the most part." 181 The accidental in this sense, it may 
cogently be reasoned, has the possibility of not existing and, by the principle that 
every possibility is eventually realized, could not be eternal. Maimonides estab- 
lished that the motion of a power attached to a body is accidental in a certain 
sense, in the sense that such a power cannot undergo motion in and by itself, or 
essentially — only bodies can — yet is moved by being associated with something 
that does undergo motion in itself and essentially. To complete Maimonides' 
argument, the sense in which the motion of the soul is accidental would have to 
be shown to be identical with, or to imply, the other sense of accidental. Motion 
that is accidental in the sense of being an accompaniment of essential motion 

179 Above, pp. 243-244, 250-251. 
180 Above, pp. 246-247. See n. 62. 

181 Metaphysics V, 30, 1025a, 14-15. Significantly, accident in a second sense, ibid., 31-33, is 
that which "belongs to each thing by virtue of itself but is not in the essence . . . ; accidents of this 
sort may be eternal." What is accidental in the second sense obviously does not have the possibility 
of not existing. 



i 



( 



272 Proof from Motion 

would have to be shown to be accidental in the sense of not being necessary; 
only then would there be cogency in the contention that the motion of a soul of 
the spheres must, because it is accidental, have the possibility of not existing and 
accordingly must eventually cease. 

Crescas states an objection along analogous lines. He writes: "What exists by 
accident has the possibility of not existing" solely in instances "where it does 
not result necessarily from something existing essentially [and eternally]." By 
contrast, "accidental things that do result necessarily from essential things" will 
"be eternal, should the latter be eternal." Now "when we ascribe accidental 
motion to a soul of a sphere," we do so only insofar as the soul is "connected to 
the sphere." That is to say, the motion of the soul is accidental only in the sense 
that a soul is moved not in and by itself, but thanks to its presence in a body; the 
motion of the soul is not accidental in the critical sense of not being necessary. 
There is consequently no reason to suppose that the motion of the soul of the 
sphere has a possibility of ceasing. 182 The soul of a celestial sphere might thus 
very well move its sphere unceasingly. Since the motion undergone by the soul 
would be the necessary concomitant of an unceasing essential motion — the motion 
of the sphere which is produced by the soul itself — it too would be unceasing 
and necessary. And we would have a situation in which the motion of the soul 
of the sphere, although accidental in a certain sense, is yet necessary and eternal, 
and hence a situation in which the soul eternally performs its task of moving the 
sphere. 

(iv) Summary of Crescas' critique of the argument for the 
incorporeality of the first mover 

To establish the incorporeality of the first mover of the universe and complete 
the proof from motion, three alternative explanations of the movement of the 
celestial spheres had to be ruled out. The first conceivable explanation of the 
movement of the spheres was the hypothesis that the mover is a body which is 
moved by another body, and so on ad infinitum. Crescas ostensibly rejects the 
argument designed to rule out this alternative, by defending the possibility of an 
infinite number of bodies. In fact, however, Crescas acknowledges that every 
series of causes can be traced back to a first cause. He thereby acknowledges that 
philosophic reasoning can establish a first mover which is moved by nothing else, 
and hence a mover that is not a body, all bodies being moved by something apart 
from themselves. 183 

The second conceivable explanation of the motion of the spheres was the 
hypothesis that the first mover of the universe is a power distributed through a 
body. Crescas contends that two — or three— indispensable premises in the argu- 
ment ruling out the hypothesis are unfounded; that even granting all the premises, 

182 0r ha-Shem, I, ii, 5; Wolfson, Crescas, pp. 248-253 and nn. 6, 8. Aristotle, Physics VIII, 6, 
259b, 28-31, seems to be saying the same. 
'"Above, p. 251. 



Proof from Motion 



273 



the argument ruling out the hypothesis remains invalid, since a power finite in 
intensity might be infinite in continuity; and that even granting the validity of the 
entire argument, the argument can still be circumvented by identifying the factor 
moving the spheres as the nature of the spheres.'" 4 

The third conceivable explanation of the eternal motion of the spheres is the 
hypothesis that the spheres are sustained in their eternal motion by a soul or souls 
attached to them. Crescas contends that the argument designed to rule out the 
hypothesis is again not cogent; for although the soul of the spheres would undergo 
motion that is accidental in a certain sense, such motion would not be accidental 
in the specific, critical sense implying nonetcrnity. 

The upshot of Crescas' critique is that Aristotelian argumentation cannot estab- 
lish the correctness of the fourth alternative, the thesis that the motion of the 
spheres and of the universe as a whole is due to an incorporeal being independent 
of the spheres. The first mover might be an inanimate power distributed through 
the body of the spheres, or the nature of the spheres, or the soul of the spheres. 
The proof from motion thus fails to demonstrate the incorporeality — or tran- 
scendence, if one will — of the first mover, and consequently fails as a genuine 
proof of the existence of God. Crescas' critique incidentally discloses the extent 
to which the argument or arguments for the incorporeality of the first mover are 
grounded in the presuppositions of Aristotelian physics. The arguments can be 
accomplished only by granting, inter alia, Aristotle's definition of body and of 
place, his theory of the physical elements and natural place, his general theory 
of motion, and his theory of natural rectilinear and natural circular motion. 

(c) The unity of the first mover 

Aristotle had established the unity of the first mover in two ways, by inferring 
the attribute of unity from the attribute of incorporeality, and by inferring the 
unity of the mover from the unity of the universe and the attendant unity of its 
motion. 185 Crescas replies explicitly to the former inference and thereby also 
suggests a reply to the latter. 

In the former Aristotle had reasoned: Since "all things that are many in number 
have matter," the first mover, having no matter cannot be many in number. 186 
Crescas starts by rejecting the conclusion on the basis of his earlier refutation of 
the argument for the incorporeality of the first mover: If the incorporeality of the 
lirst mover is not demonstrable, the unity of the first mover obviously cannot be 
derived from its incorporeality. 187 

But Crescas goes further. He contends that even if the incorporeality of the 
first mover should be granted, unity still cannot be inferred from incorporeality. 
Here, as elsewhere, his contention is rooted in positions taken by his predecessors. 

""Above, p. 270. 
""Above, p. 248. 
186 Above, n. 24. 

187 0r ha-Shem, I, ii, 15 (implied). 



274 



Proof from Motion 



Aristotle, at least in certain passages, and medieval Aristotelians universally, 
recognized a multiplicity of incorporeal beings, the incorporeal movers of the 
spheres or so-called intelligences. 188 A multiplicity of incorporeal beings could 
be countenanced only by qualifying the rule according to which objects are 
differentiated solely through their matter. 189 The qualification best known to 
Crescas and the one to which he addresses himself is the qualification formulated 
by Maimonides. As was seen earlier, Maimonides explained that incorporeal 
beings can be differentiated from each other and enumerated in a single instance, 
when they stand in the relationship of "causes and effects." 190 Since the incor- 
poreal movers of the spheres form, in Maimonides' picture of the universe, a 
causal series in which each link emanates the next, his version of the rule regard- 
ing the differentiation of beings is compatible with the multiplicity of movers of 
the spheres: The movers of the spheres, although incorporeal, arc differentiated 
from each other inasmuch as they form a series of causes and effects. And yet, 
the rule, as qualified by Maimonides, still permits the unity of the first mover to 
be inferred. For if incorporeal beings are distinguishable only when they stand 
to each other in the relation of cause and effect, there can exist no more than one 
incorporeal being that is a cause without being an effect. Maimonides' version 
of the rule for the differentiation of beings thus accommodates a multiplicity of 
incorporeal beings, while allowing, so Maimonides thought, the unity of the first 
mover to be inferred from the first mover's incorporcality. 

Crescas demurs. The qualified rule for differentiating objects is compatible, 
he objects, not merely with a multiplicity of incorporeal beings, but also with a 
multiplicity of incorporeal first movers. A multiplicity of first movers would be 
conceivable, should "one be [understood to be] the cause of one effect and another 
[understood to be] the cause of another effect." That is to say, uncaused causes 
can be differentiated from each other even though one uncaused cause obviously 
cannot stand to another in the relation of cause to effect. Uncaused causes can 
be differentiated on the supposition that each has its own proper effect. To be 
specific, should the emanationist scheme that had been grafted adventitiously on 
Aristotle's cosmology be disregarded, the movers of the several celestial spheres 
would not stand to each other in a causal relation, yet would be differentiated 
inasmuch as "one is the cause of one sphere and another is the cause of another 
sphere." Each incorporeal mover would be a first cause in its own right, the 
several first causes would be differentiated, and the universe would not depend 
on a single deity. 191 Alternatively, a multiplicity of unconnected physical uni- 
verses might conceivably exist. Aristotle had, to be sure, undertaken to demon- 
strate the impossibility of more than one universe; 192 but Crescas, in another 



188 Cf. Metaphysics XII, 8; above, n. 25. 

""Various qualifications of Ihe rule are discussed by Wolfson, "The Plurality of Immovable 
Movers," pp. 8-13. 

l90 Above, n. 70 [principle (19)]. 

""Or ha-Shem, I, ii. 15. m De Caelo 1,8-9. 



Proof from Motion 



275 



connection, answered Aristotle on that issue and defended the possibility of a 
multiplicity of universes. 193 Should several universes in fact exist, the first cause 
of each would be differentiated from the first cause of the others by virtue of its 
having a different effect. Each universe would have its own first cause, and the 
totality of universes would not depend on a single deity. 194 

The foregoing is Crescas' reply to the argument that the first mover must be 
one because it is incorporeal. The other Aristotelian argument for the unity of 
the first mover consisted in inferring the unity of the mover from the unity of the 
universe and its motion. By defending the possibility of a multiplicity of physical 
universes, Crescas suggests an appropriate reply to this argument also: If a num- 
ber of distinct universes exist, a number of first movers would likewise exist, 
each universe having its own mover. 

To summarize: Crescas maintains that the unity of the first mover cannot be 
inferred from the attribute of incorporcality since the incorporeality of the first 
mover had not been demonstrated successfully. Furthermore, even granting the 
incorporeality of the first mover, the attribute of unity still does not follow, since 
a multiplicity of incorporeal first movers might be differentiated by virtue of 
having as their effects cither different parts of a single universe or independent 
universes. In addition, Crescas' defense of the possibility of several unconnected 
universes suggests an answer to Aristotle's other argument for the unity of the 
first mover, the inference of the unity of the first mover from the unity of the 
universe. 

Crescas' analysis of the proof from motion reveals that the proof is grounded 
in Aristotle's physics, its presuppositions are the presuppositions of Aristotle's 
physics, and many of those presuppositions can be overturned from within the 
Aristotelian system. Moreover, even when all the premises in the proof are granted, 
the hoped for conclusions can still not legitimately be drawn. As a consequence, 
the proof completely fails to establish two of the three theses required for a 
genuine demonstration of the existence of God. It fails to establish that only a 
single first mover exists and it fails to establish that the first mover or movers are 
incorporeal or even animate. 

4. Another proof from motion 

The first thesis in the proof from motion is the existence of a first unmoved 
mover, 195 and the primary consideration adduced by Aristotle in arriving at a 
first mover was, as seen earlier, the impossibility of an infinite regress of causes. 6 
Aristotle adduced several secondary considerations as well, and among them is 



m Or ha-Shem, I, ii, I (4); Wolfson, Crescas. pp. 216-217. 
""Or ha-Shem, 1, ii, 15. 
I " 5 Above, pp. 241-242. 
196 Above, n. 35. 



276 



Proof from Motion 



what has been called the argument from "logical symmetry." 197 That is an argu- 
ment resting on a principle asserting: When two things are known to exist in 
combination with each other, and when in addition one of the two is known to 
exist separately, then the other must likewise be assumed to exist separately. The 
principle is illustrated by the Greek commentators through the example of honey- 
water 198 or honey-wine. 199 Honey-water or honey-wine is known to exist as a 
compound, and honey is in addition known to exist separately; therefore the 
second component in the compound must, so the thinking went, also be assumed 
to exist separately. 

The necessity of qualifying the principle was pointed out by the Greek com- 
mentators. Alexander observed that the principle is not valid in the case of com- 
pounds of substance and accident. 200 That is to say, although a given substance 
and a given accident exist together in a compound, and although, further, the 
substance may, by its nature, exist independently of accidents, 201 nevertheless 
the accident may not be concluded capable of separate existence. Simplicius too 
observed that the principle is not applicable universally and, quoting an expres- 
sion used by Aristotle in connection with it, 202 he characterized the principle as 
no more than "reasonable." An exception was discovered by Simplicius in the 
realm of speech. In speech, vowels and consonants are sometimes pronounced 
together, and in addition, vowels arc sometimes pronounced separately; never- 
theless, consonants can never be pronounced separately. 203 

The principle that when one element in a compound exists separately the other 
must likewise exist separately was, as already mentioned, adduced by Aristotle 
as a secondary consideration in the course of the proof from motion. Aristotle's 
precise argumentation remains slightly blurred because of textual uncertain- 
ties, 204 but for our purposes the way in which subsequent philosophers read him 
will suffice. Aristotle was typically understood to have argued: Sense perception 
testifies in the first place to the existence of something that both undergoes motion 
and produces motion, and in the second place, to the existence of something that 



" 7 T. Gompcrz, Greek Thinkers, Vol. IV (London, 1912), p. 219; Wolfson, "Nolcs on Proofs of 
the Existence of God," p. 573. 

'"Alexander, quoted by Averroes, Long Commentary on Metaphysics, XII, comm. 35; German 
translation in J. Frcudcnthal, Die durch Averroes erlialtenen Fragmente Alexanders ztir Metaphysik 
des Aristoteles (Berlin, 1885), pp. 107-108. 

'"Themistius, Paraphrase of Physics, ed. H. Schenkel, Commentaria in Aristolelem Graeca, Vol. 
V/2 (Berlin, 1900), p. 223; Simplicius, Commentary on Physics, p. 1227. 

200 See n. 198. Aquinas makes a similar statement, Commentary on Physics, VIII, §1044. 

20l In fact, material substances are never devoid of accidents, and therefore Alexander's point is 
weak or at least difficult to bring out. Aquinas, in the passage referred to in the previous note, writes 
that the principle of logical symmetry is not applicable to accident and substance, since "accident is 
in substance per se," whereas the contrary is not true. 

202 Aristotle, Physics VIII, 5, 256b, 23. 

203 Simplicius, Commentary on Physics, p. 1227. 

204 See Ross's apparatus and note to Physics VIII, 5, 256b, 20-24. 



Proof from Motion 



277 



undergoes motion without producing motion. Accordingly it is, in Aristotle's 
words, "reasonable if not necessary" to conclude that something also exists which 
produces motion without undergoing motion. 2 " 5 

The argument from logical symmetry establishes, at host, an uncaused cause 
of motion. But no argument for the existence of an uncaused cause would, as we 
know, reach the level of a complete proof of the existence of God until supple- 
mented by arguments for incorporeality and unity. 206 In Aristotle, the argument 
from logical symmetry did not require special supplementation, because of the 
context in which it was employed. It had served as nothing more than a subsidiary 
consideration for reaching a first mover; and once a first mover had been reached 
through it or through the primary Aristotelian consideration that causes cannot 
regress infinitely, Aristotle had completed his overall demonstration through the 
arguments for the incorporeality and unity of the first mover which were examined 
in earlier sections of the present chapter. 207 

Alexander and Themistius proposed another procedure by which a first mover 
could be shown to be incorporeal, explaining that an unmoved mover, being 
subject to no motion or change, would be free of potentiality, and hence free of 
matter. 208 Joining this other explanation of the incorporeality of the first mover 
to the argument from logical symmetry gave rise to a new, not overly profound, 
proof of the existence of God. An unmoved mover was established through the 
principle of logical symmetry; and the unmoved mover was found to be incor- 
poreal because free of potentiality. 209 The unity of the unmoved mover still 
remained to be inferred, but, as was seen in connection with the standard proof 
from motion, the attribute of unity is derivable from the attribute of incorpo- 
reality. 210 The existence of a single, incorporeal, unmoved mover— a deity — 
could thus be demonstrated. 

In the Middle Ages, the argument from logical symmetry is not frequent, but 
it is used. The corpus of Arabic works attributed to Alexander has it in two 
guises, neither of which, however, seems intended as a complete proof of the 
existence of God, their function being rather to exhibit God's attributes. The 
argument in its first guise follows Aristotle's example and establishes an unmoved 
mover. 2 " The alternate guise looks at purely actual, as distinct from potential, 



205 Aristotle, ibid. Cornford, in the Locb edition of the Physics, Vol. II, p. 336, observes that the 
passage seems out of place. The argument also appears in Aristotle, Metaphysics XII, 7, 1072a, 23- 
26. 

206 Above, pp. 3, 242. 
21,7 Above, pp. 244, 248. 

2(,8 Cf. Aristotle, Metaphysics XII, 2, 1069b, 24-25. 

209 Alcxandcr, in the passage cited in n. 198; Themistius, Paraphrase of Metaphysics, interpreting 
Metaphysics XII, 7, 1072a, 23-26. Arabic text in Arista 'ind al-'Arab, p. 14; Hebrew text in 
Commentaria in Aristolelem Graeca, Vol. V/5, ed. S. Landauer (Berlin, 1903), p. 16. 

2l0 Above, p. 248. 

211 Alexander, Mabadi' al-Kull, in Aristii 'ind al-'Arab, pp. 257-258. 



278 



Proof from Motion 



being and runs: Something is discovered to exist which is purely potential, to 
wit, prime matter; 212 and other things arc discovered to exist which are both 
potential and actual; consequently, something must exist as well which is purely 
actual. 213 

In Maimonides, the argument from logical symmetry gains the status of a full- 
fledged proof of the existence of God and is enumerated as the second of four 
such proofs, the standard proof from motion having been enumerated by him as 
the first. From "Aristotle," Maimonides adduces the "principle . . . that if some- 
thing is found composed of two things, and in addition one of those things is 
found by itself apart from the composite, then the other must also exist apart 
from the composite." As an illustration, Maimonides cites the example of "honey- 
vinegar." 214 He thereupon proceeds to apply the principle to the phenomenon of 
motion. There exist things "composed of mover and moved," that is to say, things 
producing motion while themselves undergoing motion; and things that arc "moved 
without moving anything"; whence Maimonides concludes that in addition 
"something must exist which moves without being moved," that is to say, a "first 
mover." Since what docs not undergo motion is "indivisible and incorporeal," 215 
the unmoved mover must be incorporeal. 216 Maimonides' readers are left to infer 
the unity of the first mover from the attribute of incorporcality; the reasoning 
whereby that inference can be drawn had already been spelled out by Maimonides 
in his statement of the standard proof from motion. 217 

In scholastic philosophy a version of the argument from logical symmetry is 
employed by William of Auvergne. From the presence in the universe of beings 
that both bestow and receive existence, and of other beings that only receive 
existence, William concludes that something must exist which only bestows exis- 
tence. 218 Albertus Magnus, who knew Maimonides' philosophy well, employs 
the argument from logical symmetry to establish the existence of an unmoved 
mover. And he characterizes the argument as the "strongest proof of the existence 
of God." 219 

Crescas' critique 

Maimonides had not merely cited the principle underlying the argument from 
logical symmetry. He had justified it, explaining that if the two components in 
question were such that "they can exist exclusively in combination, . . . neither 

212 0n the Aristotelian physical scheme, prime matter is never in fact found in isolation. 
21 Alexander, Mabadi al-Kull, p. 272. 
2,4 Cf. above, p. 276. 

2 "Cf. above, pp. 247-248, principle (16). 
2 ' 6 Guide, II, 1 (2). 
2n Guide, II, 1 (1). 

2I *S. Schindele, Beitraege zur Melaphysik des Wilhelm von Auvergne (Munich, 1900), pp. 47- 

49. 

2 ' 9 Albcrtus Magnus, De Causis el Processu (part of his Parva Naluralia) I, 1,7; 9-10. 



Proof from Motion 



279 



of them would under any circumstances be able to exist independently of the 
other. The fact that one of the two docs exist separately proves that there is no 
necessary interconnection between them. Consequently, the other too must exist 
separately." 220 

Crescas responds that the logic is faulty. If one of two components in a com- 
pound exists separately and accordingly is not bound to the second component, 
the latter may legitimately be judged to have the "possibility of existing" sepa- 
rately. Nothing, however, justifies concluding that the latter actually does exist 
separately. Therefore, Crescas maintains, the principle at issue, when applied to 
motion, will at most permit inferring that an unmoved mover may exist, not that 
an unmoved mover actually docs exist. 221 

A type of composite is moreover pointed out by Crescas — who is now repeat- 
ing an observation made by Gersoniclcs — where the principle is not at all appli- 
cable, where empirical evidence shows that not even the possibility of separate 
existence can be supposed. In the hierarchy of nature, each lower stage is dis- 
covered to exist independently of the one above it, whereas the higher cannot 
exist except in conjunction with the one below. Plainly, the separate existence of 
the lower stage, taken together with the combined existence of the lower and 
higher stages does not permit the conclusion that the higher too exists separately. 
For example, objects exist which possess the vegetative faculty alone, and others 
exist which possess both the vegetative and animal faculties; but nothing exists 
which possesses the animal faculty without the vegetative faculty. 222 Crescas, 
following Gcrsonidcs, is suggesting that the attempted inference of an unmoved 
mover falls precisely in the area where the principle does not apply. The char- 
acteristic of only undergoing motion is at a lower stage than the characteristic of 
only producing motion. If the principle at issue docs not permit inferring the 
separate existence of a higher stage from the separate existence of a lower stage, 
the principle docs not permit inferring the existence of something that produces 
motion without undergoing motion from the existence of things both undergoing 
and producing motion, taken together with the separate existence of things only 
undergoing motion. 223 



""Guide. II, 1 (2). 

2zl 0r Iw-Shem, I, ii, 16. For the criticism lo hold, Crescas would have to explain why the rule 
that every possibility is eventually realized docs not apply. 

222 Ibid. Sec Gcrsonidcs, Milhamal ha-Shem (Leipzig. 1866), V, iii, 6. p. 267 Gcrsonidcs is 
refuting the use of the argument from logical symmetry to establish the existence of (he movers of 
the spheres and not strictly its use as a proof of the existence of God. 

22, Crcscas offers a further criticism, which appears differently in the two editions and in the 
manuscript at my disposal. I was not able to reach a satisfactory interpretation of any of the versions. 

The argument from logical symmetry might also be challenged by maintaining (hat nothing at 
all docs exist which is moved without moving something else. For objects that seem to be moved 
without moving anything else do, in fact, move the air around them; and each portion of air moves 
the portion beyond it. 



280 



Proof from Motion 



In fine, the underlying principle of the argument from logical symmetry, which 
the Greek commentators already insisted must be restricted, 224 is rejected by 
Crescas on both logical and empirical grounds. And Crescas finds that the argu- 
ment from logical symmetry, like the standard proof from motion, fails as a 
demonstration of the existence of God. The standard proof failed to establish the 
incorporeality and unity of the first mover. The argument from logical symmetry 
fails to establish even the existence of a first unmoved mover. 



Above, p. 276. 



IX 



Avicenna's Proof of the Existence 
of a Being Necessarily Existent 
by Virtue of Itself 



1. First cause of motion and first cause of existence 

Aristotle's proof from motion, ostensibly at least, sought to establish a first 
incorporeal cause solely of the motion of the universe, not of the existence of the 
universe. For its existence, the universe would, in Aristotle's system, be self- 
sufficient, depending on no further cause outside itself. Plotinus subsequently 
expounded the conception of a cause of the very existence of the universe, 1 and 
later Ncoplatonist philosophers draw the contrast between a first cause of existence 
and a first cause of motion. They also introduce the possibility of a cosmological 
proof establishing a cause of the former sort. 

We find the contrast between a first cause of existence and a first cause of 
motion in Proclus, who draws it in order to distinguish Plato's position from the 
position of Aristotle and his followers. Plato, according to Proclus, had main- 
tained that the entire universe must depend for its existence on an "efficient" 
cause, whereas Aristotle and his school failed to recognize an efficient cause of 
the existence of the universe. Nevertheless, in Proclus' view, Aristotle and the 
Peripatetics would have approached the correct Platonic position had they only 
seen the full significance of the proof from motion. Central to the proof from 
motion is the principle that since every body must be finite, it can contain only 
finite power. Employing the principle, Aristotle inferred that the power contained 
in the heavens is insufficient to maintain eternal motion; and inasmuch as the 
heavens do move eternally, they must depend for their motion on an incorporeal 
cause outside themselves. 2 That was as far as Aristotle went. But, Proclus con- 
tends, the same principle also applies to existence. The principle shows that the 

'Cf. E. Zcllcr, Die Philosophic der Griechen. Vol. III. Part 2 (5th. ed.: Leipzig. 1923). pp. 530 
ff., 550 ff. 

2 Cf. above, p. 244. 



281 



282 



Avicenna's Proof 



power contained in the heavens is insufficient to maintain eternal existence as 
well as eternal motion. And inasmuch as the heavens do exist eternally, they must 
depend for their existence, and not merely Cor their motion, on an incorporeal 
cause outside themselves. In the reasoning of Aristotle's proof from motion Pro- 
clus thus detects the germ of a proof for an incorporeal cause of the very existence 
of the corporeal, though not of the incorporeal, 3 realm. 4 

Simplicius — following, as he says, a book by his teacher Ammonius, 5 who, 
it happens, had at one time been a student of Proclus — states the issue in a similar 
fashion, but with one addition. Simplicius too observes that on the reading of 
such peripatetics as Alexander of Aphrodisias, Aristotle established a "final and 
motive cause of the heavens, not an efficient cause [of their existence). " He further 
contends that the principle that a body can contain only finite power is "clearly" 
not limited to "motive power." It applies as well to the power "constituting the 
being" of a corporeal object, and "therefore [the heavens] . . . must receive even 
their eternal corporeal being from an incorporeal [cause)." 6 So far Simplicius has 
not diverged from the account given by Proclus. But Simplicius had undertaken 
a general harmonization of Aristotle with Plato. Accordingly, he goes beyond 
Proclus and contends that not Aristotle, but his students, failed to see the impli- 
cations of the proof from motion. Aristotle himself "was in agreement with his 
master [Plato] in holding God to be ... an efficient cause of the [existence of 
the) entire universe and the heavens," and Aristotle intended his proof from 
motion to establish that thesis. 7 To support his reading of Aristotle, Simplicius 
has recourse to a subtle, rather unconvincing, exegesis of the Aristotelian text. 1 * 

In late Greek Neoplatonic philosophy we find, then, the view that Aristotle's 
proof from motion implied a proof of a first efficient cause of the very existence 
of the physical universe, although there was a difference of opinion as to whether 
Aristotle himself was conscious of the implication. The tradition apparently fil- 
tered down to Avicenna's time. In a commentary on Aristotle's Metaphysics XII, 
Avicenna criticizes the proof from motion on the grounds that it establishes no 
more than a first cause of motion; it is "incapable ... of establishing the one, 
the True, who is the first cause (mabcla') of all existence." 9 Yet, despite his 

■'The principle that a finite body cannot contain infinite power would not apply to incorporeal 
beings. See, however, below, n. 37. 

"Proclus, Commentary on Timaeus, cd. E. Diehl, Vol. 1 (Leipzig, 1903), pp. 266, 295. Also 
cf. the passage from a lost work of Proclus quoted by John Philoponus, De Aeierniniie Muntli contra 
Proclum, cd. H. Rabc (Leipzig, 1899), pp. 238-239. 

5 Thc book was known to Alfarabi. Cf. M. Mahdi, "Alfarabi against Philoponus," Journal of 
Near Eastern Studies, XXVI (1967), 236, n. 9. 

''Simplicius, Commentary on the Physics, cd. H. Dicls, Commentaria in Aristotclem Gracca, 
Vol. X (Berlin, 1895), pp. 1362-1363. 

7 Ibid., p. 1360. 

8 lbid., pp. 1361-1362. 

'Avicenna, Commentary on Metaphysics XII, in Arista 'ind al-'Arah. ed. A. Badawi (Cairo, 
1947), p. 23. 



Avicenna's Proof 283 

criticism of Aristotle's proof, Avicenna disputes the position of certain "com- 
mentators," including one Abu Bishr, according to whom Aristotle recognized 
nothing more than a first cause of motion. Avicenna cites a statement of Aristotle's 
in the Metaphysics to the effect that "the heavens depend" on the "first cause 
(mabda')"; and he explains the statement as an acknowledgment that the first 
cause of motion is also a first cause of the very "existence (qiwcimY of the 
universe. 10 Avicenna, in other words, understands that although the proof from 
motion docs not establish a first cause of existence, Aristotle did recognize that 
a complete picture of the universe requires construing the first cause of motion 
as a cause of existence as well. 

Elsewhere, in another collection of philosophic notes, Avicenna adds a refine- a 
mcnt. He takes up the following question: If a proper proof of the existence of' 
God must establish a "first cause (mabda') of the existence and substance [of the • 
universe)," why did the earlier philosophers limit themselves to proving a "cause 
merely of motion?" Avicenna's brief reply is that "openly," indeed, the earlier 
philosophers demonstrated merely a first cause of motion; yet by "allusion" and 
by "implication (bi-l-quwa)" they also proved a first cause of the existence of 
the universe." The present passage still assumes that the proof from motion is * 
unable in itself to establish anything more than a first cause of motion. It adds >. 
that Aristotle had not only acknowledged a first cause of existence, but had 
consciously alluded to and implied — perhaps, but not necessarily, in his proof 
from motion 12 — another proof which is able to establish a first cause of the very- 
existence of the universe. 

Avcrrocs too found a proof of a first cause of existence implied in Aristotle. v 
Aristotle's proof from motion, Avcrrocs explains, is tantamount to a proof of a> 
first cause of existence. He writes: "Should the motion [of the heavens) cease, / 
the heavens [themselves] would cease"; for the essential nature of the heavens is ✓ 
such that they must move, and without motion they could not possess their cssen- < 
tial nature as heavens. Furthermore, "should the movement of the heavens cease, 
the movement of what exists under the heavens would likewise cease." 13 Without 
the first cause of motion, "the universe as a whole would cease to exist" — at 
least in the form of a universe. 14 Aristotle's first cause of motion is consequently 
a first cause of the existence of the universe. 15 

'"ibid., p. 26, referring to Metaphysics XII, 7, 1072b, 14. An identification of the Abu Bishr 
whom Avicenna mentions is made by H. Brown, "Avicenna and the Christian Philosophers in Bagh- 
dad," Islamic Philosophy and the Classical Tradition (Walzer Festschrift), (Oxford and Columbia, 
South Carolina, 1972). pp. 43-45. 

"Avicenna, Muhahatlu'it, 5290, in Aristu 'ind al-'Arah, p. 180. Also cf. $264, p. 173. 

1 J Thc medieval philosophers found another proof of the existence of God in Metaphysics II; sec 
below. Chapter XI. 

L1 Cf. above, p. 242. 

,4 It could still presumably exist as inert matter. 

'H)e Substantia Orbis. IV; cf. Aristotle, De Caelo I, 9, 279a. 28-30. 



284 



Avicenna's Proof 



2. The existence of God: a problem for metaphysics 

Avicenna's proof of a first cause of the existence of the universe is characterized 
by him as a metaphysical proof, anil Aviccnna attached considerable significance 
to that characterization. The notion is elaborated most fully in Avicenna's most 
comprehensive philosophic work, the Shifa'. There Avicenna undertakes to define 
the "subject matter" (mawdu') of the science of metaphysics and in particular to 
show that the subject matter of metaphysics is not the "deity (innJya allah)." 16 
He employs three Aristotelian rules regarding the subject matter of any science, 
and finally arrives at Aristotle's position on what the proper subject matter of 
metaphysics is. 17 And in the course of his discussion he shows that one task of 
metaphysics — as distinguished from the subject matter of metaphysics — is the 
demonstration of the existence of God. 

Aristotle had explained that a science (a) does not "demonstrate" the existence 
of its own subject matter. Astronomy and mathematics, for example, do not 
"demonstrate" the existence of celestial and mathematical entities, but rather 
"accept" their existence. Nor (b) docs a science demonstrate its "principles," the 
most fundamental premises from which it reasons; these too it accepts. The 
procedure of each science is (c) to "examine" various "essential attributes," whose 
"meaning" is accepted but whose "existence" remains to be proved, and to 
demonstrate that the attributes do belong to the subject matter. Arithmetic, for 
example, accepts the existence of its subject matter, number, and proceeds — with 
the aid of the principles of arithmetic, which it also accepts, together with the 
aid of subordinate premises, which it derives from the principles — to demonstrate 
the essential attributes of number. 18 

Using these rules, Avicenna shows that the deity cannot be the proper subject 
matter of the science of metaphysics. His reasoning is as follows: The existence 
of God — this Avicenna posits with no explanation — is surely neither "self-evi- 
dent" nor "unprovable"; 19 it must, hence, be "amenable to proof." Since no 
science apart from metaphysics could possibly prove the existence of God, meta- 
physics remains the only place were the proof can be undertaken. By Aristotle's 
rules, however, the existence of the subject matter is not to be demonstrated by 
a given science, but is rather to be "accepted." Consequently, metaphysics, which 
docs have the task of demonstrating the existence of the deity, must have some- 
thing other than the deity as its subject matter. Put in another way, since the 
existence of God is a "subject of inquiry" (matlub) of the science of metaphysics, 
God cannot be the "subject matter" (mawdu') of metaphysics. 20 ^ 

"'Shifa: llahiyat, ed. G. Anawati and S. Zayed (Cairo, I960), p. 5; French translation, with 
pagination of the Arabic indicated: La Mitaphysique du Shifa', trans. G. Anawati (Paris, 1978). 

l7 Thc assumption is that the subject matter of a given science is not arbitrary, but something to 
be determined objectively. 

18 Posterior Analytics I, 9-10. 

"Cf. F. Rosenthal, Knowledge Triumphant (Leiden, 1970), p. 21 1. 
20 Shifa' : llahiyat, p. 6. 



Avicenna's Proof 



285 



Avicenna was cognizant of an objection that would immediately occur to stu- 
dents of Aristotle. The existence of a first mover hail been proved in greater detail 
in Aristotle's Physics than in his Metaphysics. How then might it be maintained 
that demonstrating the existence of God is a task for the science of metaphysics 
and not for the science of physics? In answering the objection, implied rather 
than expressly stated by him, Avicenna points out that physics is the science of 
corporeal entities, whereas the first cause is known even by physics to be incor- 
poreal. The existence of the first cause cannot therefore be a proper subject of 
inquiry for physics, and when physics docs treat of the first cause, it impinges 
upon an issue not genuinely its own. Physics merely "gives a fleeting idea" of 
the existence of God, thereby encouraging philosophers to pursue the subject in 
the discipline proper to it, that is, in the science of metaphysics. 21 

Having established that the existence of the deity is a subject of inquiry of 
metaphysics, and thereby having ruled out the deity as the proper subject matter 
of metaphysics, Avicenna follows Aristotle in stating just what the subject matter 
of metaphysics is. Aristotle had explained that metaphysics "examines the exis- 
tent qua existent and what belongs to it by virtue of itself." 22 Avicenna accord- 
ingly concludes, after some additional discussion, that the subject matter, whose * 
existence is simply accepted by the science of metaphysics, is "the existent qua / 
existent." The subject of inquiry of metaphysics is, stated broadly, the attributes ? 
that "belong to the existent merely by virtue of its being existent." 23 In other- 
words, the task of metaphysics is to demonstrate that certain attributes do indeed , 
belong to the existent solely insofar as it is existent. 

Here Avicenna is led to consider a further difficulty. Aristotle had laid down/\J 
the rule that a science does not demonstrate its own principles, but only accepts 
them. 24 Avicenna construes the rule as meaning both that a science does not 
demonstrate the principles from which it reasons — that is to say, principles in 
the sense of primary premises — and also that it does not demonstrate the existence 
of the principles of its subject matter — that is to say, principles in the sense of 
causes of existence. 25 When applied to the science of metaphysics, the rule shows 
that the "principles" of what exists cannot be part of the subject of inquiry of 
metaphysics; for were that the case, metaphysics would improperly be demon- 
strating the existence of the principles of its own subject matter. The subject of 



21 Ibid., pp. 6-7. Alfarabi similarly states that metaphysics has the task of demonstrating the 
existence of incorporeal beings; cf. Alfarabi, Ihsa al-'Ulum, ed. and trans. A. Gonzalez Palcncia 
(Madrid, 1953). Arabic part, p. 88. 

"Aristotle, Metaphysics IV, 1. 1003a, 20-21, and D. Ross's note in his edition (Oxford. 1924). 
Cf. Alfarabi, Aghrad ma ha'd al-TabVa, in Rasa il al-Farabi (Hyderabad, 1931), p. 4; below, p. 
313. 

"Shifa : llahiyat, p. 13. 24 Posterior Analytics I, 10. 

25 In the passage cited in the previous note. Aristotle appears to be making only the first of these 
two points. But both senses of principle arc Aristotelian: cf. Metaphysics VI, 1 , 1021b. 34 If. Averroes 
understood Aristotle's rule in the way Avicenna did; cf. below, p. 314. 



286 



Avicenna's Proof 



inquiry of metaphysics must, as just seen, be the "attributes belonging to" the 
existent insofar as it is existent. 26 These considerations lead to a two-sided dif- 
ficulty, which is taken up by Avicenna: First, the deity seems to be a principle or / 
cause of everything that exists, and hence a principle of the subject matter of , 
metaphysics; by Aristotle's rule, the existence of God would therefore have to , 
be accepted, and not demonstrated, by metaphysics. Secondly, in any event, the . 
existence of the deity hardly seems to be an attribute belonging to the existent , 
qua existent, and therefore hardly an appropriate subject of inquiry for metaphysics. > 
,> In resolving the first side of the twofold difficulty, Avicenna notes that God 
cannot after all be a principle or cause of the entire subject matter of metaphysics. 
The subject matter of metaphysics is the existent insofar as it is existent, "the 
existent with no further qualification." And God cannot be the cause of the "exis- 
tent without qualification," since he clearly is not the cause of absolutely every- 
thing that exists, not being a cause of himself. 27 God is the cause of only a portion 
of existence, the portion that is "caused." Inasmuch as God is the principle of/ 
only part of the subject matter of metaphysics, it is, Avicenna contends, legitimate / 
for metaphysics to undertake a demonstration of his existence; for it is legitimate > 
for a science to demonstrate the existence of the principle of part of its own > 
subject matter. 

t As for the second side of the objection raised by Avicenna — according to which 
the existence of the deity, if indeed a subject of inquiry of metaphysics, would 
have to be an attribute belonging to the subject matter — that ceases to be odd on 
closer inspection. The proof of the existence of God establishes that the property 
of not being constituted by anything else is exemplified in the realm of actual 
existence. In a sense, therefore, the proof of the existence of God docs demon- 
strate that a certain attribute is applicable to the existent qua existent; 28 and the 
proof of the existence of God turns out to be a proper subject of inquiry for the 
science of metaphysics. 29 

Avicenna concludes, then, that the subject matter of the science of metaphysics 
is the "existent qua existent." The subject of inquiry of metaphysics can, he 
writes, be broken down into three subheadings: the attributes of the existent qua ^ 
existent; the ultimate causes of caused existents, the investigation of which, as ,( 
just shown, can be construed as part of the inquiry into the attributes of existence; ^ 
and also the principles of the individual sciences. 30 /- 



26 Cf. above, p. 285. 

27 Avicenna will undertake to prove that God exists by reason of himself, but would reject as 
senseless the description of God as cause of himself. 

28 The proof does not, of course, establish that the property in question is applicable to all 
existence. The case of attributes such as unity and plurality would be analogous. Metaphysics studies 
the applicability of those attributes and similar pairs to the existent insofar as it is existent, even 
though neither member of the pair belongs to all existence; cf. Aristotle, Metaphysics V. 

2 "Shifa: llahiyat. p. 14. 

"'Ibid., pp. 14-15. Cf. Alfarabi, IhstV al-'Ulum. pp. X7-89. On the third subheading, cf. 11. 



Avicenna's Proof 



287 



The foregoing discussion of the subject matter and subject of inquiry of the 
science of metaphysics, with its justification of a metaphysical proof of the exis- 
tence of God, is found in Avicenna's Shifa'. Avicenna touches upon the question 
of the metaphysical proof of the existence of God in the Isharat as well. 

There he discusses the method that he pursues in his proof of the existence of 
God. The proof, he writes, consists in "examining nothing but existence itself"; 
by "considering . . . the nature {lull) of existence," the proof has "existence qua 
existence testify to the first [cause]." 31 The method thus delineated is contrasted 
by Avicenna with another, whereby the existence of God is established not from 
a consideration of existence in general, but instead from a consideration of one 
segment of existence, namely "creation and effect." Although the latter method, 
which takes its departure from "creation and effect," is recognized by Avicenna 
as valid, his own method, he insists, is "more certain and more exalted." 32 

The difference between the two methods is stated here in language that is 
deliberately allusive but easily deciphered. Metaphysics, as was seen, is the 
science that "examines the existent qua existent and what belongs to it by virtue 
of itself." When Avicenna claims to have constructed a proof exclusively by 
examining "existence itself," "existence qua existence," and "the nature of exis- 
tence," he plainly means that he has constructed a proof using propositions drawn 
exclusively from the science of metaphysics. He contrasts his metaphysical proof 
with the proof that begins with God's "creation and effect" and that reasons back 
from them to the existence of God as a first cause. Avicenna's intent cannot be 
that the metaphysical proof uses absolutely no data drawn from God's "creation 
and effect." For, it will appear, his proof does require at least one datum from 
the external world; 33 and the parts of the world accessible to man arc man himself 
and physical nature, both of which belong to the realm of "creation and effect." 
Avicenna's meaning must be that the metaphysical proof considers no properties 
peculiar to the realm of "creation" but instead considers the attributes of objects 
belonging to that realm solely insofar as they are existent. As for the proof of the 
existence of God which is not a metaphysical proof hut focuses on God's "creation 
and effect," it can be nothing other than a physical proof of the existence of God, 
that is to say, a proof based not on the attributes of existence in general, but on 
the attributes peculiar to physical existence. The metaphysical proof, Avicenna 



Wolfson. "The Classification of Sciences in Medieval Jewish Philosophy," reprinted in his Studies in 
the History of Philosophy and Religion. Vol. I (Cambridge, Mass.. 1973), p. 5 IK; and D. Ross's 
introduction to his edition of Aristotle's Prior and Posterior Analytics (Oxford. 1949). p. 64. 

"The sentence continues: "whereupon he testifies concerning everything in existence posterior 
to him." This is a Sufi theme; cf. e.g. , Ibn 'Ata' Illah, K. al-Hikam, trans. V. Danner (Leiden. 1973), 
§5 29, 249. 

,2 K. al-lsharat wa-l-Tanbihat, cd. J. Forget (Leiden. 1892), p. 146. The French translation by 
A. Goichon. I.ivre des Directives el Remarques (Beirut and Paris. 195 1 ), gives the pages of Forgets 
edition. 

"Cf. below, p. 303. 



288 



Avicenna's Proof 



asserted, is "more certain and more exalted" than the other proof, but the asser- 
tion is not supported by an explanation as to where the superiority of the meta- 
physical proof lies. Scattered statements elsewhere do, however, furnish an expla- 
nation. 

Avicenna has written that since physics is the science of corporeal entities, a 
physical proof can give no more than a "fleeting idea" of the existence of an 
incorporeal first cause. 34 He has also put aside Aristotle's proof from motion on 
the grounds that it is incapable of "establishing the . . . first cause of all exis- 
tence." 35 In a similar vein, he stresses that the metaphysical proof has the advan- 
tage of establishing a cause of "every caused existent . . . insofar as it is caused" 
and "not merely insofar as it is moved or quantitative." 36 These statements indi- 
cate at least two virtues of the metaphysical proof: It establishes a cause of the i 
entire universe, incorporeal as well as corporeal, and it establishes a cause of the \ 
very existence of the universe. A physical proof, by contrast can at most establish j, 
a first cause only of the corporeal realm and, in the instance of Aristotle's well-/ 
known proof from motion, it establishes a first cause exclusively of the motion/ 
of the corporeal realm. 37 

It is easy to go beyond Avicenna's explicit statements and expand upon the A 
virtues that he could have perceived in a metaphysical proof, rendering it superior < 
to Aristotle's proof from motion. The proof from motion rested on a set of physical 1 
principles: Motion in place underlies all other kinds of change; 38 everything* 
moved has the cause of its motion outside itself; 39 nothing can maintain itself in< 
motion unless it is continuously moved by an agent; 40 only circular motion is< 
continuous; 41 only an infinite force can maintain the heavens in motion for an' 
infinite time; 42 the heavens cannot contain an infinite force. 43 Using all the phys-/ 
ical principles, Aristotle, undertook to establish the existence of an unmoved 
incorporeal cause solely of the motion of the universe. Avicenna, although not 
rejecting Aristotle's physical principles, dispenses with them in his metaphysical 
proof. And yet, without them, he is confident that he can prove the existence of 
an incorporeal first cause not merely for the motion of the physical universe — 
that being the only part of the universe in motion — but for the very existence of 
the entire universe. The metaphysical proof proves more with fewer premises. It 
travels, or attempts to travel, further with less fuel. 



34 Cf. above, n. 21. 35 Cf. above, n. 9. 

M Shifa': Ilahiyal, p. 14. 

37 An argument for the unity of God could, however, show that all incorporeal beings outside the 
first cause are also dependent on the first cause for their existence. Cf. above, p. 274. 
38 Cf. above, p. 241. 
39 Cf. above, pp. 242-247. 
40 Cf. above, p. 242. 
41 Cf. above, p. 242. 

42 Cf. above, p. 244, n. 49; pp. 245-246. 
43 Cf. above, pp. 244-245. 



Avicenna's Proof 



289 



3. Necessarily existent being and possibly existent being 

The metaphysical proof of the existence of God to be examined in the present 
chapter appears in two of Avicenna's works, at length in the Najat and somewhat 
obscurely in the Isharat. 44 In both works, Avicenna analyzes the concepts nec- 
essarily existent being and possibly existent being and uses the analysis to estab- 
lish the existence of a being necessarily existent by virtue of itself. The concepts 
necessarily existent being and possibly existent being arc also analyzed by Avi- 
cenna in two other works, the Shifa' and Danesh Niimeh 45, but there they serve 
to define the nature of God, not to establish his existence, and the existence of 
God is established in a different manner. 46 In discussing Avicenna's analysis of 
the concepts necessarily existent and possibly existent I draw upon all of the 
works just mentioned. But in discussing the particular proof of the existence of 
God which is the subject of the present chapter I limit myself to the Najat and 
! Isharat/ 1 ' 

The proof of the existence of God as a necessarily existent being probably was 
suggested to Avicenna by a passage in Aristotle. In the Metaphysics, Aristotle 
presented a version of his proof from motion, whereupon he added a postscript: 
Since the prime mover "can in no way be otherwise than as it is," it "is an existent 
. . . of necessity" — an existent of necessity from which "the heavens and nature 
depend Avicenna's proof, particularly the fuller version in the Najat, can be 
understood as starting where Aristotle left off. Avicenna sets aside all the physical 
arguments leading up to Aristotle's first cause that is an "existent ... of neces- 
sity." lie begins afresh by analyzing the concept "existent of necessity" or nec- 
essarily existent, as he calls it, working out everything contained in the concept 
as he construes it. Then, adducing a single datum from the external world, he 
undertakes to establish that something necessarily existent by virtue of itself does 
in fact exist. 

Before, however, entering into any analysis of metaphysical concepts, includ- 
ing necessarily existent and possibly existent, Avicenna points out that primary^ 
concepts cannot strictly be defined at all. Definitions in Aristotelian logic are x 
framed by taking a wider and already known concept, the genus, and setting/ 
apart a segment of it through a specific difference. It follows that primary con- < 
ccpts, which are not "subsumed under anything better known" and hence arc part > 
of no genus, cannot be defined. They are rather "imprinted in the soul in a primary ' 
fashion," 49 and must be grasped immediately. The thesis that primary concepts 

"Najat (Cairo. 1938), pp. 224 ff.; K. al-lsharal, pp. 140 ff. 

"'Shifa : Ilahiyal. pp. 37 ff., 343 ff.; Danesh Nameh (Teheran. 1937), pp. 105 ff.; French 
translation: Le Livre de Science, transl. M. Achena and H. Masse (Paris, 1955). pp. 136 If.; English 
translation; The Metaphysica of Avicenna. trans. P. Morewcdge (New York, 1973), pp. 18 ff. 

""Below, Chapter XI. 

47 Thc most important passages in the Najat and Shifa are translated by G. Hourani. "Ibn SIna 
on Necessary and Possible Existence," Philosophical Forum, IV (1972), 74-86. 

""Aristotle, Metaphysics XII, 7, 1072b. 10-14. 49 Shifa': IMhixat, p. 29. 



290 



Avicenna's Proof 



cannot be defined is developed most comprehensively by Avicenna in connection 
with the general concepts existent and thing, and is thereupon applied to the , 
concepts necessary, possible, and impossible; Ihose concepts too, he writes, , 
cannot be "made known ... in a true sense." 5 " 

Ostensible definitions of necessary, possible, and impossible had been offered, 
notably by Aristotle, but Avicenna contends that the ostensible definitions lead 
in a vicious circle. He considers two ostensible definitions of necessary: "that 
which can (yumkin) not be assumed [to be] absent (ma'dum)"; 51 and that which 
is such that an impossibility would result if it should be assumed to be other than 
it is." 52 The first of the two definitions employs the concept possible (mumkin) — 
"can" (yumkin) — and the second employs impossible. When we turn to ostensible 
definitions of possible we find that they in their turn employ either the concept 
necessary or impossible. For possible is defined as "that which is'not necessary; 
or which is absent (ma'dum), but is such that its existence is not impossible if it 
should be assumed to occur at any time in the future." 53 And ostensible definitions 
of impossible, for their part, employ either necessary or possible. 54 Attempts to 
define the triad thus chase one another in a circle. 55 Nevertheless, although 
primary concepts are not explicable by anything wider and better known, and are 
consequently inaccessible to true definition, there is, Avicenna understands, a 
way of presenting them to the man who for some reason does not have them 
imprinted in his soul. One may "direct attention" to the primary notions and "call 
them to mind" through a "term or an indication." 56 On that basis, Avicenna 
ventures an explication of necessary: "It signifies certainty of existence." 57 

The distinction between necessary and possible is employed by Avicenna to 
distinguish two types of being. A "necessarily existent being" is a being that 
"perforce exists"; alternatively, it is "such that when it is assumed not to exist, 
an impossibility results." A "possibly existence being" is a being that "contains 
no necessity ... for either its existence or nonexistence ('adam)"; alternatively 
it is "such that whether assumed not to exist or to exist, no impossibility results." 58 

50 Ibid., p. 35. 

"Implied in Aristotle, Prior Analytics I, 13, 32a, 19-20. 
"Cf. Aristotle, Metaphysics V, 5, 1015a, 34. 
"Cf. Aristotle, Prior Analytics I, 13, 32a, 19-20. 

54 The impossible is "what is necessarily nonexistent," or "what cannot (la yumkin) exist." Cf. 
Aristotle, De Interpretation 13, 22a, 19-20; 22b, 6. 
"Shifa': llahiyat, p. 35. 
56 Ibid., p. 29. 
"Ibid., p. 36. 

is Najat. pp. 224-225. Avicennae Metaphysices Compendium, trans. N. Carame (Rome, 1926), 
p. 66, is misleading. Carame's translation reads: "Neccsse-essc est ens quod si ponatur non esse 
implicat contradictionem (mahal). Possible vero esse est illud quod sivc ponatur esse sive non esse, 
non inde oritur repugnantia (mahal)" The mistranslation of mahal as "contradictio" suggests that 
denying the existence of a necessarily existent being involves a logical contradiction. Cf. below, p. 
394. 



Avicenna's Proof 



291 



What Avicenna is offering is clearly not strict definitions. Necessarily existent » 
can hardly be defined, in the strict sense, through impossibility, since impossibility > 
could only be defined through necessity or through possibility, which in its turn - 
would have to be defined through one of the other two terms. What Avicenna 
has given are explications of the type that merely "direct attention" to the meaning 
of concepts and "call them to mind." 59 

The distinction between necessarily existent and possibly existent is supple- 
mented by Avicenna with a further distinction, suggested by Aristotle. When 
Aristotle analyzed the term necessary, he observed that "for certain things, some- 
thing else is a cause of their being necessary." But "for some, nothing is |a cause 
of their being necessary]; rather it is through them that others exist of necessity." 60 
That is to say, there is a class of things necessary without having a cause of their 
being so; and a second class of things necessary through a cause, a cause to be ' 
found in the former class. The distinction drawn by Aristotle in regard to nec- 
essary suggests to Avicenna a distinction between two types of necessarily exis- 
tent being. One can, Avicenna writes, conceive of a being as necessarily existent 
either by virtue of itself or by virtue of something else. The former would be 
something "such that because of itself and not because of anything else what- ' 
soever, an impossibility follows from assuming its nonexistence." The latter would 
be a being "such that it becomes necessarily existent, should something other 
than itself be assumed | to exist]." The illustrations Avicenna adduces for the latter 
category arc "combustion," which is "necessarily existent . . . when contact is 
assumed to take place between fire and inflammable material," and "lour," which 
is "necessarily existent . . . when we assume two plus two [to exist]."''' Now if 
anything is necessarily existent by virtue of something else, it must — since it 
will not exist by reason of itself without that other thing — be "possibly existent 
by virtue of itself." 62 In all, Avicenna thus envisages three categories: (a) the , 
necessarily existent by virtue of itself; (b) the necessarily existent by virtue of > 
another, but possibly existent by virtue of itself; and (c) the possibly existent by * 
virtue of itself which is not rendered necessarily existent by virtue of another. , 

Here Avicenna makes a point that has escaped some students of his philos- 
ophy. 61 Of the three categories, the first two, he states unambiguously, are the 
only conceivable categories of actual existence. Everything actually existent, 
including everything "entering existence" in the physical world — as combustion, 
to take the illustration used by Avicenna — is necessary in one sense or the other. 64 
Put in another way, the possibly existent can never actually exist unless rendered 



59 Above, n 56. Aristotle, Metaphysics V. 5. 

'''Najat, p. 225 For Avicenna, it must be stressed, "2 plus 2 equals 4" is not logically necessary. 
"Najat. p. 225. 

63 As, for example, Averrocs; cf. below, p. 318. Some modern scholars have also missed Avi- 
cenna's point. 

M Cf. Shifa : lluhiyat, p. 37. 



292 



Avicenna's Proof 



necessary by something distinct from itself. 65 Aviccnna\s division of being hereby 
differs from, for example, Alfarabi's. Alfarabi applied the designation possibly 
existent to those objects that actually exist, yet have the possibility of not existing 
and are hence unable to exist forever. In other words, he designated all actual 
transient objects in the sublunar world as possibly existent with no further qual- 
ification; and he restricted the designation necessarily existent to beings that 
cannot cease to exist, that is, to eternal beings. 66 Avicenna, by contrast, insists 
that all objects that actually exist, even transient beings, are to be characterized 
as necessarily existent, and all objects that exist by reason of something else, 
even eternal beings, are possibly existent. Both sets belong to the single category 
of the necessarily existent by virtue of another, possibly existent by virtue of itself. 
Alfarabi's usage is unquestionably more genuinely Aristotelian than Avicenna's. 67 

In justifying his designation of all actually existent beings as necessarily exis- 
tent, Avicenna employs a train of thought similar to that underlying arguments 
from the concept of particularization. 68 The possibly existent can enter the realm 
of actual existence, he reasons, only if a factor distinct from itself should "dif- 
ferentiate out" existence for it in preference to nonexistence. And once a factor 
of the sort is present, the possibly existent being perforce exists; for its existence 
is rendered necessary. 69 The proper way of construing possible existence, accord- 
ing to Avicenna, is therefore to say that during the time the possibly existent 
actually exists, its existence is necessary, and during the time it docs not exist, 
its existence is impossible. The necessity and the impossibility of its existence 
are both conditioned, due not to itself, but to the presence or absence of an 
external condition, which necessitates its existence or nonexistence. Considered 
in itself, in isolation from the external condition— and only considered in that 
way — the possibly existent at all times remains possible. 70 

Actual existence is then either: (a) Necessarily existent by virtue of itself; this 
is something "such that if assumed not to exist an impossibility results," with the 
proviso that it has its character by reason of itself. Or (b) necessarily existent by 
virtue of another, but possibly existent by virtue of itself; this is something, 



65 Cf. Najat, p. 226. 

'"'Alfarabi, al-Siyasdt al-Madaniya (Hyderabad, 1927), pp. 26-37. There is a suggestion of 
Avicenna's usage in Alfarabi, Commentary on Aristotle's De Interpretalione, ed. W. Kutsch and S. 
Marrow (Beirut, 1960), p. 192. 

67 For Aristotle, necessary and eternal arc mutually implicative; cf. below, pp. 294, 319. 

68 Cf. above, pp. 161 f.; 178f. Averroes saw the Kalam influence on Avicenna in this point. Cf. 
Averroes, K. al-Kashf, cd. M. Mueller (Munich, 1859), p. 39; German translation, with pagination 
of the original Arabic indicated: Philosophie and Theologie von Averroes, trans. M. Mueller (Munich, 
1875); Long Commentary on Physics, in Aristotelis Opera cum Averrois Commentariis, Vol. IV 
(Venice, 1562), II, comm. 22; Tahafut al-Tahafut, ed. M. Bouyges (Beirut, 1930), p. 276; English 
translation with pagination of the original Arabic indicated: Averroes' Tahafut al-Tahafut, trans. S. 
van den Bergh (London, 1954). 

69 Thus all events in the universe occur necessarily. 

10 Shifa': llahiyat, pp. 38-39; Najat, pp. 226, 238. 



Avicenna's Proof 



293 



again, such that if assumed not to exist, an impossibility results, with the proviso 
that it has its character by reason of another, only inasmuch as "something other 
than itself is assumed [to exist]." 71 The necessity characterizing the two categories 
of necessarily existent being, is, as already seen, construed by Avicenna as an 
indefinable primary concept to be grasped by the human mind immediately. 72 As 
a mere "indication" of the meaning of necessity Avicenna wrote that the term 
"signifies certainty of existence." 73 The necessarily existent by virtue of itself 
would accordingly be that which has certainty of existence by virtue of itself; 
the necessarily existent by virtue of another would be that which has certainty 
of existence by reason of another. And the impossibility involved in assuming 
the nonexistence of a necessary being would not be any logical impossibility, but 
would consist in contradicting the certainty of its existence, the fact that it does 
exist. 74 If no more can be said about the meaning of necessarily existent, it is 
difficult to see how necessary existence differs from actual existence. Indeed, 
necessary existence for Avicenna seems simply to be actual existence, with the 
added understanding that whatever actually exists, exists by necessity, and with 
the further understanding that necessity is a primary concept, the meaning of 
which must be grasped immediately. 

So far, it must be stressed, Avicenna's analysis has been conducted exclusively 
in the realm of concepts, and he has not committed himself to the existence of 
anything. 75 lie has merely stated that whatever might be assumed to exist would 
have to be classified as cither necessarily existent by virtue of itself or necessarily 
existent by virtue of another. 

4. The attributes of the necessarily existent by virtue of itself 

Having established that whatever actually exists is either necessarily existent 
by virtue of itself or necessarily existent by virtue of another, Avicenna proceeds 
to analyze the former concept and to set forth its "properties." The analysis of 
the concept is designed to serve a double function in his proof of the existence 
of God. It contributes to the argument showing that something necessarily existent 
by virtue of itself does exist, 76 and in addition it reveals that the entity in question 
possesses the attributes of a deity. Avicenna's analysis has its sources in Aristotle, 
Plolinus, Proclus, and Alfarabi, although the pertinent passages in these writers 
are not connected with a proof of the existence of God. 

Aristotle had written that the most fundamental sense of the term necessary is 
"that which cannot be otherwise." And, he continued, anything of the sort "does 

7, Above, n. 61. 

72 Above, pp. 289-290. 

"Above, n. 57. 

74 This is especially clear in the Danesh Nameh, p. 106; French translation, p. 136; English 
translation, p. 48. 

"This is clear throughout, and is stated explicitly in ShifiT : llahiyat, p. 37. 
76 Cf. below, §5. 



294 



Avicenna's Proof 



not admit of more states than one," and must therefore by "simple." Aristotle 
further suggested that what cannot be otherwise must be "eternal and immov- 
able." 77 We thus have three attributes belonging to that which is necessary in the 
sense of not being able to be otherwise: It must be simple, eternal, and immovable. 

The Neoplatonic passages that lie behind Avicenna's analysis of the concept 
necessarily existent by virtue of itself were mediated though the Liber de Causis, 
a medieval Arabic work falsely attributed to Aristotle, but in fact a paraphrase 
of Proclus' Elements of Theology. In the Liber de Causis, the "first cause" is 
described as the only entity that is "self-sufficient" (mustaghniya bi-nafsilid), 
and analysis of self-sufficiency reveals that the first cause is "simple in the highest 
degree." The reasoning runs as follows: When an object is not simple but "com- 
posite," it "stands in need of something outside itself, or [at least in need] of the 
elements of which it is composed"; that is to say, a composite entity is dependent 
upon whatever external factor joins its parts together and, even supposing that 
no such external factor, is required, a composite entity considered as a whole is 
distinct from its parts and dependent on those parts. If the composite is inescap- 
ably dependent upon something distinct from itself and hence not self-sufficient, 
anything self-sufficient cannot be composite but must be "simple." 78 The first 
cause, which is known to be self-sufficient, must therefore be simple. 

Besides analyzing the implications of the attribute self-sufficiency, the Uber 
de Causis also analyzes the implications of self-subsistence, which ostensibly has 
the same meaning. 79 If a "substance" is "self-subsistent" (qd'im bi-dhdtihi), the 
analysis now goes, it cannot "be brought into existence (mukawwan)." For were 
it brought into existence, it "would stand in need" of whatever external factor 
brings it into existence and perfects it; and it would consequently not after all be 

11 Metaphysics V, 5, 1015b, 12-15. 

™Liber de Causis, ed. and trans. O. BarHenhcwer (Freiburg, 1882), §20, and cf. §27. This 
reflects Proclus, Elements of Theology, ed. and trans. E. Dodds (Oxford, 1963), §127, which in turn 
goes back (o Plotinus, Enneads, II, 9, 1. Cf. also Dodds' note to Elements of Theology, §§9, 10. 

79 In the original Greek, self-sufficient and self-subsistent are treated as synonymous terms; cf. 
Dodds' commentary on Proclus, Elements, p. 224. In the Greek, the entities characterized as self- 
sufficient and self-subsistent stand at a level of existence below the first cause of the universe; Proclus 
held the strange view that the self-sufficient and the self-subsistent represents a level of existence 
that both is its own cause and yet also has a cause above it. Cf. ibid., pp. 196, 224. (Aviccnna, Najat, 
p. 225, refutes the thesis that something can be both necessarily existent by virtue of itself and also 
necessarily existent by virtue of another.) In the Arabic paraphrase, by contrast, the designation self- 
sufficient is explicitly restricted to the first cause. And it is debatable whether self-subsistent is 
synonymous with self-sufficient and thus also a designation of the first cause, or whether the self- 
subsistent represents a lower stage in the hierarchy of existence. §20 taken together with §27 suggests 
the former interpretation, whereas §§24, 25, 28, suggest the latter interpretation. (§20 appears to be 
a conscious effort by the author of the paraphrase to make the theory more monotheistic than was 
intended in the original Greek. For that tendency in other Arabic paraphrases and translations of 
Neoplatonic works, see G. Endress, Proclus Arabus [Beirut, 1973], pp. 206-219, 236-237, 240- 
241.) These questions of interpretation do not affect my present purpose, which is to show how 
certain attributes were analyzed out of the concepts self-sufficient and self-subsistent. 



Avicenna's Proof 



295 



self-subsistent. 1 "' The "self-subsistent" cannot, moreover, be "subject to destruc- 
tion." For things are destroyed only by being separated from the cause of their 
existence, whereas the self-subsistent can never become separated from its cause 
since it is its own cause. 81 Analysis of the concepts self-sufficient and self-sub- 
sistent has thus established that whatever is so described cannot be composite, 
cannot be brought into existence, and cannot be destroyed. And the Liber de 
Causis has employed a significant thought that was to be applied more system- 
atically by Alfarabi and Aviccnna: A composite entity taken as a whole is different 
from its components and dependent on them for its existence. 

Alfarabi, in his turn, undertakes to determine the attributes of "the First," that 
is, the first being in the hierarchy of existence and the first cause. The First, he 
contends, can, by definition, "have no cause through which, from which, or for 
the sake of which" it exists. For if it did have a cause, it would not itself be the 
first cause. 82 Not having a cause, the First must be "sufficient in itself for its own 
duration and continued existence." Now anything containing the possibility of 
being destroyed is plainly not sufficient in itself for its own duration and con- 
tinued existence. It follows that the First docs not have any possibility of being 
destroyed and hence is eternal. 83 

Furthermore, since the First has no cause, it can in no way be "divisible." To 
establish this, Alfarabi deploys an argument similar to that wherein Proclus found 
the self-sufficient to be simple because the composite "stands in need of" its own 
parts. Alfarabi reasons: The components through which a divisible, composite 
object "receives its substance" (tajawhara) arc "causes of the existence" of the 
object; and since the First can have no cause, it can have no components what- 
soever. It must consequently be completely indivisible. It must be free not only 
of "quantitative" divisibility, but also of the internal composition represented by 
the parts of a definition and the composition resulting from the joining of matter 
and form in a corporeal object. The First must be absolutely indivisible, indefin- 
able, incorporeal, and uncxtended. 84 

There can, Alfarabi continues, be no more than one First existent. If there were 
two, they would each have to possess a common clement by virtue of which they 
both deserve the common designation First; and at least one of the two would 
have to have an added element peculiar to itself by virtue of which the two entities 
could be distinguished from one another and enumerated as two. At least one of 
the two would, in other words, be composite. But, if composite, its parts, by the 
already familiar argument, would be the cause of its existence. Inasmuch as at 

""Liber tie Causis. §24; cf. Proclus, Elements. §45. 
"'l.iher tie Causis. §25; cf. Proclus. Elements, §46. 

82 Alfarabi, K. Ara AM al-Matlina al-Fudila, ed. F. Dieterici (Leiden. 1895). p. 5: German 
translation, with pagination of the original Arabic indicated; Der Mitsterstaat. trans. F. Dieterici 
(Leiden, 1900). 

K1 lbid., p. 8. and cf. p. 5. 

R4 Ibid., pp. 8-9; al-Siyasiit. pp. 14-15. 



296 



Avicenna's Proof 



least one of the two would have a cause and not after all be First, only one First 
being is conceivable. 85 

Alfarabi derives additional attributes from the concept of the First. Whatever 
is immaterial consists in pure intellect; therefore the First, being immaterial, must 
be pure actual intellect. 86 Truth designates the degree of existence a thing has; 
the First, which ex hypothesi occupies the highest degree of existence, is accord- 
ingly truth par excellence. 87 Beauty similarly is understood by Alfarabi as pro- 
portional to the perfection of existence a thing has; the First, having the highest 
degree of existence, will accordingly be of the highest beauty. 88 Pleasure consists 
in perception of beautiful objects; since the First has the most perfect perception 
of the most perfect beauty, namely himself, he enjoys the highest conceivable 
pleasure. 89 And since the First is the object of his own love and desire, he is the 
"prime object of love" and the "prime object of desire." 90 By analyzing the 
concept Alfarabi has concluded that the First must be uncaused, eternal, indivi- 
sible and simple, undefinabie, incorporeal, unextended, one, pure intellect, truth, 
most beautiful, the prime object of love, and possessed of the highest pleasure. 91 

Avicenna's analysis of the concept necessarily existent by virtue of itself runs 
along the same lines. The necessarily existent by virtue of itself, he contends, 
clearly can "not have a cause." For if it did, "its existence would be by virtue of 
[that cause]" and not by virtue of itself. 92 Like Alfarabi, Avicenna gives the 
proposition the broadest application, explaining that the necessarily existent by 
virtue of itself can have a cause in no sense whatsoever; it cannot even have 
internal causes, "principles which combine together and in which the necessarily 
existent consists." The full argument for rejecting internal components of any 
kind rests on the distinction between a given entity as a whole and the parts of 
which it is composed. Any composite entity, Avicenna submits, exists by virtue 
of its parts and not be virtue of itself as distinct from the parts. As a consequence, 
it exists, considered as a whole, not through itself, but through something else — 
through the components that constitute it. It is therefore not necessarily existent 
by virtue of itself; and the necessarily existent by virtue of itself cannot, con- 
versely, be composite. 93 

The implications of the noncomposite nature of the necessarily existent by 
virtue of itself are drawn by Avicenna almost exactly as Alfarabi drew the impli- 
cations of the noncomposite nature of, the First. If the necessarily existent by 



"K. Ara' AM al-Madina, p. 6; al-Siyasat, p. 14. 
8< X. Ara' AM al-Madina, p. 9; al-Siyasat, p. 15. 
"K. Ara' AM al-Madina, p. 10. 
88 A:. Ara' AM al-Madina, p. 13; al-Siyasat, p. 16. 
>9 K. Ara AM al-Madina, p. 14; al-Siyasat, p. 16. 
^K. Ara' AM al-Madina, p. 15; al-Siyasat, p. 17. 

"This type of reasoning is also suggested in a text attributed to Alexander of Aphrodisias; cf. 
Aristu "md al-'Arab, ed. A. Badawi (Cairo, 1947) p. 266. 
"Shi/a' : llahiyat, pp. 37-38. 
"Najat, pp. 227-228. 



Avicenna's Proof 



297 



virtue of itself can contain no parts whatsoever, it is simple in every conceivable 
manner. It is incorporeal, inasmuch as it is not composed of matter and form. It 
is unextended and immaterial, inasmuch as it is free of quantitative parts. It is 
undefinabie, inasmuch as it is not composed of genus and specific difference. 
And it is free of the distinction of essence and existence. 94 

There can, moreover, be only one entity necessarily existent by virtue of itself. 
To prove the thesis, Avicenna examines, in some detail, the various ways in 
which things having a common characteristic can nevertheless be distinguished 
from each other. Basically, though, his contention is that positing two entities 
both of which arc necessarily existent by virtue of themselves would imply that 
those entities have a cause and, again, that they arc composite. The reasoning is: 
If the intrinsic nature of the "species" necessarily existent by virtue of itself is 
such as to belong exclusively to a certain given entity, it cannot belong to anything 
else. If, by contrast, the intrinsic nature of the species necessarily existent by 
virtue of itself is not such as to belong exclusively to a certain given entity, some 
added factor, for example, a substratum, must be responsible for the presence of 
the species in whatever entities happen to have it. On the latter alternative, 
however, the added factor would be the "cause" of the presence of the species, 
whereas the necessarily existent by virtue of itself has, ex hypothesi, no cause. 
The intrinsic nature of the species necessarily existent by virtue of itself must 
therefore be such as to belong exclusively to a certain entity; and only one entity 
of the sort is conceivable. 95 Furthermore, positing two entities both necessarily 
existent by virtue of themselves would amount to positing two beings that are 
similar in one respect, their necessary existence, but different in another, the 
respect whereby they can be distinguished and enumerated as two. But that 
situation would be conceivable only if at least one of the two should be composite, 
containing both the clement it has in common with its counterpart and another 
clement whereby it can be distinguished and by reason of which two distinct 
beings can be enumerated. At least one of the two would, therefore, be composite 
and, as already seen, 96 not necessarily existent by virtue or itself. 97 It follows, 
then, both from the uncaused and the noncomposite character of the necessarily 
existent by virtue of itself that not more than one is conceivable. 

Avicenna derives other attributes from the concept necessarily existent by virtue 
of itself. The necessarily existent by virtue of itself must be pure intellect, since 
beings free of matter are pure intellect. It must be true; for truth consists in the 
highest grade of existence, and the necessarily existent by virtue of itself would 
have the highest grade of existence. 98 It must be good, for evil consists in pri- 
vation, whereas the necessarily existent by virtue of itself has fullness of being 



l Shifa : llahiyat, pp. 344-348; Najat. pp. 228-229; K. al-Isharal. p. 144. 
'Najat. pp. 229-230. Cf. Shifa' : llahiyat. p. 349. 
'Above, p. 296. 

'Shifa: llahiyat. pp. 43-47. 350-354; Najat. pp. 230-234; K. al-lsharal. p. 143. 
'Cf. Aristotle, Metaphysics II, 1, 993b, 26-31. 



298 



Avicenna's Proof 



and suffers no privation. It must constitute the highest beauty, be the highest 
object of desire, be possessed of the greatest pleasure, and so forth." 

Avicenna's analysis of the concept necessarily existent by virtue of itself thus 
establishes that anything corresponding to the concept must be uncaused, simple, 
incorporeal, one, pure intellect, truth, good, most beautiful, an object of desire, 
and possessed of the greatest pleasure. The analysis mirrors Alfarabi's analysis 
of the First so consistently that we may be certain Aviccnna borrowed Alfarabi's 
analysis of the First, elaborated it somewhat, and applied it to the concept nec- 
essarily existent by virtue of itself. The critical thought is, as in Proclus and 
Alfarabi, that any composite considered as a whole is distinct from its parts and 
dependent on them for its existence; and therefore, what exists solely by reason 
of itself can have no parts. Significantly, both Proclus and Alfarabi came to their 
results without reference to necessity, and in Avicenna too the element of necessity 
plays no role. Avicenna in effect considers the implications of a thing's existing 
by virtue of itself. He could as well have analyzed and read out the attributes of 
the actually existent by reason of itself instead of the necessarily existent by 
virtue of itself. 

5. Proof of the existence of the necessarily existent by virtue of itself 

Avicenna does not regard the analysis of the concept necessarily existent by 
virtue of itself as sufficient to establish the actual existence of anything in the 
external world. He does not, in other words, wish to offer an a priori or onto- 
logical proof of the existence of God, but rather a new form of the cosmological 
proof. 

He is careful to define the degree and mode in which the existence of God can 
be established by philosophy. The existence of God is taken by him to be neither 
self-evident nor unprovable. 100 Nor can the existence of God be established 
through a syllogistic "demonstration" (burhan). 101 A truly demonstrative syllog- 
ism must be framed with propositions that arc "prior to," and the "causes" of, 
the conclusion. 102 It is, more precisely, a syllogism in which the middle term is 
the cause of the presence of the major term in the minor term. 103 Since there is 
nothing prior to, and the cause of, the presence of actual existence in the nec- 
essarily existent by virtue of itself, a demonstrative syllogism leading to the 

"Shifa : llahiydt, pp. 355-356, 367-370; Najal, pp. 229, 245. Cf. Aristotle, Metaphysics XII, 
7, 1072a, 34-35. 

'""Cf. above, p. 284. 

wi Shifa ' : llahiyal, p. 348. This was a commonplace. Cf . Alexander, Commentary on Metaphysics. 
ed. M. Hayduck, Commentaria in Aristotelem Graeca, Vol. I (Berlin, 1891), p. 686; Aporiai, cd. I. 
Bruns, Commentaria in Aristotelem Graeca, Supplementary Vol. 1 1/2 (Berlin, 1892), p. 4; Thcmis- 
tius. Paraphrase of Metaphysics, ed. S. Landauer, Commentaria in Aristotelem Graeca, Vol. V/5 
(Berlin, 1903), Hebrew part, p. 1 1; Proclus, Liber De Causis, §5. 

102 Aristotle, Prior Analytics I, 2, 71b, 19-32. 

10J Aristollc, Posterior Analytics 11, 2; Avicenna, Najal, p. 67; K. al-Ishardl, p. 84. 



Avicenna's Proof 



299 



existence of an entity of that description is impossible. What can, however, be 
provided, according to Avicenna, is a "proof" (dalil). 104 A "proof" is a syllogism 
wherein the middle term is the effect rather than the cause of the presence of the 
major term in the minor term; it is a chain of reasoning that moves not from the 
prior to the posterior, but from the posterior to the prior, from the presence of 
the effect to the existence of the cause. Stated in another way. a strictly demon- 
strative syllogism establishes both "that" a certain proposition is true and "why" 
it is true, whereas a "proof" establishes only "that" it is true. 105 A proof of the 
existence of God. as distinct from a strict demonstration, will, we arc therefore 
to understand, reason from the existence of a possibly existent being to the exis- 
tence of necessarily existent being, even though the former is the effect, not the 
cause, of the latter. 106 

As Avicenna constructs his proof, it requires three philosophic principles, each 
of which he also undertakes to prove. These are (a) the principle of causality; (b) 
the impossibility of an infinite linear regress of causes— two principles that are 
fundamental to most cosmological proofs of the existence of God; 1 " 7 and (c) the 
impossibility of a circular regress of causes. Avicenna ingeniously establishes the 
three principles through an analysis of the concepts possibly existent by virtue of 
itself and necessarily existent by virtue of itself. Significantly, the second and 
third principles arc not genuinely needed for his proof; Avicenna has. without 
quite realizing il, developed a cosmological proof (hat can dispense with the 
impossibility ol an inlinilc regress. 

(a) In formulating his version of the principle of causality, Avicenna employs 
a distinction between the cause of the "generation" (liuduth) of an object and the 
cause of its "maintenance" (tluibdt) in existence. 108 The cause of generation is 
more obvious, since no one, Avicenna is certain, can doubt that whenever an 
object comes into existence, it does so by virtue of something else. 109 But Avi- 
cenna could not pursue a first cause of the generation of every possibly existent 
being, since he believed that some possible beings arc eternal and have no cause 
of generation. Furthermore, by establishing a first maintaining cause he will 

'"'Shifa: llahiyal, p. 6. 

105 Aristotle. Posterior Analytics I. 6, 75a, 33-35, and Ross"s note; 1, 13. 78a. 22 ff.; Avicenna. 
Najal, p. 67; Shift!' : Burhan, ed. A. Aflili (Cairo, 1956), pp. 79-80. Also cf. Encyclopedia of 
Islam, (2nd cd.; Leiden, I960-), s.v. dalil, 1. van Ess, Erkenntnislehre des 'Adudaddm al-lci (Wies- 
baden, 1966), p. 367. 

""'Staled in scholastic terminology, the existence of God is susceptible to a demons/ratio quia 
although not to a demonstratio propter quid. Cf., e.g., Aquinas. Summa Theolopae. I. 2. 2, resp. 

l07 Cf. H. Wolfson, "Notes on Proofs of the Existence of God in Jewish Philosophy." reprinted in 
his Studies in the History of Philosophy and Religion, Vol. I, p. 572. 

'""[-"or the conception of a maintaining cause, cf. De Mundo VI; Plotinus, Lnneads, VI. 4. 2; 
Theology of Aristotle, ed. [•'. Dictcrici (Leipzig. 1882), p. 79; English translation: I'lotini Opera, cd. 
P. Henry and H. Schwyzer, Vol. II (Paris and Brussels, 1959), p. 245; Proclus, Elements, §13, and 
Dodds' notes. 

""Najal, p. 236. 



300 



Avicenna's Proof 



establish not merely a first cause that exercised its causality at a moment in the 
past and withdrew, but, as it were, a stronger deity, a first cause that continually 
maintains the universe in existence. Avicenna's proof herein parallels Aristotle's 
proof from motion. The proof from motion similarly sought not a cause that 
initiates the motion of the spheres, the motion of the spheres being eternal for 
Aristotle, but a cause that continually moves the spheres, 110 in other words, a 
maintaining cause of the motion of the spheres. 

Avicenna gives his attention, then, to a maintaining cause— not to the cause 
maintaining a moving object in motion, but, more comprehensively, to the cause 
maintaining a possibly existent object in existence. He looks at objects of the 
type he had designated as possibly existent by virtue of themselves, necessarily 
existent by virtue of another, that is to say, objects that actually exist although 
they arc in themselves only possibly existent." 1 Concerning any such object, 
Avicenna reasons, irrespective of whether it is generated or eternal, we may 
legitimately ask what maintains it in existence. Since the possibly existent is 
something that by definition does not exist by virtue of itself, should a possibly 
existent object actually exist, some factor distinct from it would have to be respon- 
sible for its existence. And some factor would have to be present and maintain 
the possible object in existence as long as the object exists; for even when the 
possibly existent is already actual it never ceases to be possible by virtue of itself 
and thereby dependent on something else for its continued existence. 112 Avicenna 
acknowledges that the maintaining factor may be a component within the total 
object. For example, the factor maintaining a statute in a given form is the stability 
of the material from which the statute is made. 1 13 But a component is still different 
from the whole" 4 so that here too, the factor maintaining the object is distinct 
from the object considered as a whole. If the component — for instance, the 
stability of the material — is also possibly existent, inquiry can, of course, legit- 
imately be made regarding the factor maintaining it in existence. 

The analysis of the concept possibly existent by virtue of itself— to be precise, 
merely asking what possibly existent means— has disclosed that if anything pos- 
sibly existent should exist, it must at all times depend on a cause distinct from 
itself to maintain it in existence. 

(b) The second premise established by Avicenna is formulated by him as the 
impossibility that "causes go to infinity," the impossibility of an infinite regress 
of causes. In fact, unlike other philosophers, and unlike his own procedure in 
another work, 1 15 Avicenna does not, in the proof we are now examining, directly 
argue that an infinite regress is, for one reason or another, absurd. Instead he 

"°Cf. above, p. 238. 
"'Cf. above, pp. 291-292. 
" 2 Cf. above, p. 292. 
'"Najat. p. 237. 
" 4 Cf. above, p. 294. 
" 5 Cf. below, p. 339. 



Avicenna's Proof 



301 



first argues for the broader proposition that the totality of actually existent possible 
beings, "whether finite or infinite," 1 16 must depend on a being necessarily existent 
by virtue of itself; and then, from the broader proposition, he infers the impos- 
sibility of an infinite regress as a corollary. 

Avicenna is thinking of a situation wherein z, for example, is maintained in 
existence by v, which exists simultaneously with it; and wherein y is maintained 
in existence by x, which likewise exists simultaneously; ad infinitum. To show 
that a situation of the sort cannot account for the totality of existence, he mentally 
collects into a single group all possible beings actually existing at a single moment, ' 17 
asks what maintains the group in existence, examines all conceivable answers to 
the question, and arrives at the alternative he deems correct. His reasoning is: 
The totality of possibly existent beings, taken as a whole, must be cither (a) 
necessarily existent by virtue of itself or (B) possibly existent by virtue of itself. 
The former alternative would imply the thesis that the "necessarily existent [by 
virtue of itself] is composed of possibly existent beings," which would be "a 
contradiction." Just where the contradiction lies is not made explicit by Avicenna. 
But granting the conception of necessarily existent by virtue of itself as that which 
docs not even have internal factors making it what it is, the contradiction is plain: 
If something does not have even internal factors making it what it is, it cannot, 
as Avicenna earlier pointed out, be composite." 8 

Since the totality of possibly existent beings existing together at any moment 
cannot (a) constitute an entity that is necessarily existent by virtue of itself, there 
remains (3) the second alternative according to which the totality, taken as a 
whole, is possible by virtue of itself. But the possibly existent by virtue of itself 
needs something to maintain it in existence. Hence, on this alternative, "whether 
the group is finite or infinite," it stands in need of a factor that will continually 
"provide [it] with existence." The factor, Avicenna assumes, must be cither (Bl) 
within the group or (B2) outside it. Assuming that the whole group is (Bl) 
ultimately maintained by one of its own members would, however, be tantamount 
to assuming that the member in question is a cause of itself. For to be a cause of 
the existence of a group is "primarily" to be the cause of the individual members; 
and inasmuch as the supposed cause is itself one of the members, it would perforce 
be a cause of itself. Yet the supposed cause has already been assumed, as one of 
the members of the group, to be possibly existent; and the possibly existent is 
precisely what does not exist by reason of itself. Therefore it could not be the 
cause of the collection of which it is one member. 

If the totality of possibly existent beings cannot (a) form a group that is 
necessarily existent by virtue of itself, and if, moreover, the ultimate maintaining 

1 "*Cf. the passage from Aristotle cited below, p. 337. 

" 7 llc has no qualms about treating an infinite number of objects as a single totality. 

""Ghazali, Tahiifut al-Falusifa, cd. M. Bouygcs (Beirut, 1927), IV, §6, understood Avicenna in 
this way; English translation in Averroes' Tahaful al-Tahafut, trans. S. van den Bcrgh. (London, 
1954), p. 161. 



302 



Avicenna's Proof 



cause cannot be (Bl) one of its own members, the sole remaining alternative is 
that what does maintain the totality of possibly existent beings in existence is 
((32) something outside the group. And, since, by hypothesis, all possibly existent 
beings were included inside, anything left outside is not possibly existent, but 
must be necessarily existent by virtue of itself. 

Avicenna should have stopped here. The totality of all actual beings that are 
possibly existent by virtue of themselves, he has concluded, depends on a being 
that is necessarily existent by virtue of itself; and that is as much as is required 
for his proof of the existence of God. But Avicenna wanted an explicit statement 
of the impossibility of an infinite regress of causes. Once he has established that 
a linear series, constituted by the possible beings existing together at any moment, 
must depend on something necessarily existent by virtue of itself, he goes on to 
infer, as a kind of corollary, that the scries must also be finite. The scries of 
possibly existent causes must, he reasons, "meet" its necessarily existent first 
cause and "terminate" there; and as a consequence, an infinite regress of causes 
is impossible — a regress, it may be repeated, of the type in which all the links 
exist together and each link maintains the next in existence." 9 

(c) Avicenna understands that his proof requires one more principle, the impos- 
sibility of a self-contained regress of causes, a regress that is "circular" and 
"finite" rather than infinite and linear. A circular regress is a situation wherein 
x, y, and z, for example, exist simultaneously in the manner that x is the cause 
maintaining y in existence, y is the case of z, but z is the cause of x. The 
impossibility of a situation of the sort can, Avicenna explains, be exhibited in 
two ways. First, the impossibility of a self-contained circular regress can be 
exhibited by a "similar proof" to that whereby, in establishing his second prin- 
ciple, an infinite linear regress of possibly existent beings was shown to be 
impossible. 120 Avicenna means that all the links in the supposed circular chain, 
like the links in a linear chain, would be caused and possibly existent by virtue 
of themselves; taken as a totality, the chain would remain possibly existent; an 
additional factor would be required in order to render it necessary and actual; 
and since the series would have to meet and terminate at the additional factor, it 
could not after all form a closed circle. Secondly, a self-contained circular regress 
is shown to be absurd by an argument applying only to it. In the circular regress 
x y z, x would be a distant cause of z, and z would be the immediate cause of x. 
x would consequently be a distant cause of itself, which Avicenna regards as 
absurd. By the same token, x would be a distant effect of itself, which is equally 
absurd. And the point can be made again in a slightly different way, as follows: 
x would be dependent for its existence upon something — z — whose existence is 
posterior to it. But "when the existence of something depends upon the existence 
of something else that is essentially posterior to the first, the existence of the first 

n9 Najal, p. 235; K. al-lsharal, pp. 141-142. 
""Najat, p. 236. 



Avicenna's Proof 



303 



is impossible." 121 A self-contained circular regress of causes cannot, therefore, 
exist. 

In addition to these three philosophical principles — the principle of causality, 
the impossibility of an infinite linear regress of causes, and the impossibility of 
a finite circular regress of causes — Avicenna leaves the conceptual realm for a 
single empirical datum in order to accomplish his proof of the existence of God: 
"There is no doubt that something exists (anna huna mtjiidan)." 122 It makes no 
difference what happens to exist or what the object's peculiar properties might 
be; for the purpose of his proof Avicenna is concerned with the "existent qua 
existent" 123 and therefore all he needs is the fact that something does in truth 
exist. Applying the proposition that there arc only two conceivable categories of 
actually existing beings, 124 Avicenna proceeds: "Everything that exists is cither 
necessary [by virtue of itself], or possible [by virtue of itself and necessary by 
virtue of another]. On the former assumption, a necessarily existent [by virtue 
of itself] has been established, and that was the object of our proof; on the other 
assumption, we must show that the existence of the possible [by virtue of itself 
but necessary by virtue of another] ends at the necessarily existent |by virtue of 
itself]." 125 That is to say, if the random existent object with which we start is 
conceded to be necessarily existent by virtue of itself, it may be assigned all the 
attributes already shown to belong to such a being, and the proof of the existence 
of God is complete. But the real issue is of course posed by the other alternative, 
the assumption that the random object with which we start is necessarily existent 
only by virtue of another, and possibly existent by virtue of itself. A true proof 
of the existence of God has the task of showing that anything possibly existent 
by virtue of itself ultimately depends for its actual existence upon something 
necessary by virtue of itself. 

Assuming that the actually existent object we start with is possible by virtue 
of itself, it must be maintained in existence by something else that exists together 
with it (principle of causality). The other factor, in turn, must be cither necessary 
by virtue of itself or possible by virtue of itself. If it is assumed to be necessary 
by virtue of itself, the proof is again at once complete. If, on the contrary, it is 
assumed to be possible by virtue of itself, it too must depend on a further factor 
distinct from it and existing together with it. Once again, Avicenna asks whether 
the new factor is necessary by virtue of itself or possible by virtue of itself. It is 
inconceivable, he has established, that anything should be maintained in existence 

'"Najal, p. 236; (his principle is not brought into the statement of the proof in K. al-hharat. pp. 
141-142. 

Cf. Aristotle, Physics VII] , 5, 257b, 13-20; Alexander!?), Mabadi' al-Kull, in Arislii 'ind al- 
•Amb, p. 259. 
' 22 Nojiil. p. 235. 
121 Above, pp. 286-287. 
124 Above, p. 291. 
'"Najal, p. 235. 



304 



Avicenna's Proof 



cither by an infinite linear regress of causes or by a circular regress of causes. 
The series of causes maintaining any given thing in existence must consequently 
terminate at a being that is necessarily existent by virtue of itself. 126 And the 
latter may now be assigned all the attributes earlier shown to belong to the 
necessarily existent by virtue of itself: It is uncaused, simple, and incorporeal; 
there is only one being answering the description; it consists in pure intellect; it 
is, in the highest degree, true, good, beautiful, an object of desire, and possessed 
of pleasure. 127 

The syllogism that encapsules the entire "proof (dalil) 12 * might be summa- 
rized thus: Possibly existent beings are traceable to a being necessarily existent 
by virtue of itself (from the three philosophic principles established by Avi- 
cenna). 129 Something exists which is, presumably, possibly existent by virtue of 
itself (empirical datum). 130 Therefore something exists which is traceable to a 
being necessarily existent by virtue of itself; and the latter also exists. 

6. Questions raised by Avicenna's proof 

Underlying all of Avicenna's argumentation is his analysis of the concept nec- 
essarily existent by virtue of itself, an analysis dependent on an unusual turn of 
thought. Avicenna contends that whatever truly exists by reason of itself cannot 
exist by reason even of internal factors making it what it is. As a consequence, 
the necessarily existent by virtue of itself cannot contain internal factors and must 
be free of all composition; and carrying the reasoning forward shows that there 
can exist no more than one such being. 131 The analysis is essential to the entire 
proof because it helps establish the existence of a being necessarily existent by 
virtue of itself 132 and also enables Avicenna to assign to the being in question 
the attributes of a deity. 133 At first blush Avicenna appears to be speculating about 
the correct meaning of a concept known to him through an immediate intuition 
of some kind. 134 But if that is in fact all he is doing, one can demur at his intuition 
and affirm, on the contrary, that a being may be deemed necessarily existent by 
virtue of itself even when existing by virtue of its parts; and the entire proof 
constructed upon the analysis of the concept would collapse. Avicenna's proce- 
dure can be defended only if it is something other than the taking of one side in 
a purely verbal or intuitive dispute, a dispute regarding whether the necessarily 
existent by virtue of itself may or may not properly be said to exist by virtue of 



' 26 Najat, p. 239; K. al-Isharat, pp. 141-142. 

l27 Cf. above, pp. 297-298. 

l28 Cf. above, p. 299. 

'"Above, pp. 299-302. 

,30 Above, p. 303. 

"'Above, p. 294. 

'"Above, pp. 301, 302. 

'"Above, pp. 297-298. 

134 Cf. above, p. 290. 



Avicenna's Proof 



305 



its parts. His procedure can be defended only if understood not as seeking to 
discover the meaning of the term but rather as working from a definition. 135 He 
must be understood to have arbitrarily defined necessarily existent by virtue of 
itself as that which, taken as a whole, exists solely by virtue of itself, not by 
virtue even of internal factors making it what it is; and to have analyzed out the 
implications of the definition. 

The heart of Avicenna's proof is the train of reasoning whereby he establishes 
that possibly existent beings ultimately owe their existence to a being necessarily 
existent by virtue of itself. 136 As a first step, he rules out the thesis (a) that all 
possibly existent beings existing at any one moment might, as a totality, comprise 
a being that is necessary by virtue of itself; in effect, what is being ruled out is 
that the physical universe as a whole could exist necessarily by virtue of itself. 
Avicenna does not explain exactly why possibly existent beings cannot add up to 
a being necessarily existent by virtue of itself; but, as was pointed out, the reason 
must be that the latter, in Avicenna's usage, can contain no components what- 
soever. 137 The properties of the necessarily existent by virtue of itself, it was just 
seen, arc nothing other than the implications of a definition. It is, consequently, 
because Avicenna has defined necessarily existent by virtue of itself in a certain 
fashion that a multiplicity of possibly existent beings cannot add up to such a 
being. 

The step ruling out alternative (a) may be read, then, as definitional. The next 
step runs: If the totality of possibly existent beings docs not comprise a being 
that is necessarily existent by virtue of itself, the totality must be (p) possibly 
existent by virtue of itself. And since the totality is possibly existent, something 
must maintain it in existence. Avicenna lays down two alternatives, namely that 
the totality is maintained either (pi) by one of the possibly existent beings 
contained within it, or (02) by something outside; he rejects the former alter- 
native and thereby affirms the latter. 138 Curiously, however, he does not consider 
a further alternative, which we may call ((33), the thesis that the totality is 
maintained in existence not be a single component but by all the components 
together. On this alternative the totality of possibly existent beings — in effect, 
the entire universe — would indeed be possibly existent in Avicenna's sense; for 
it would, taken as a whole, exist by reason of something different from itself. 
Still, it would not exist by reason of anything external to it, but would be possibly 
existent only inasmuch as it exists by reason of its own components. It would be 
possibly existent by virtue of itself, necessarily existent by virtue of its components. 

If thus challenged, Avicenna might perhaps have replied that each of the pos- 
sibly existent components would be part of the cause of the existence of the 

'■"Not a strict logical definition; cf. above, p. 291. 
'"'Above, pp. 301-302. 
'"Above, p. 301. 
""Above, ibid. 



( 



306 Avicenna's Proof 



f totality; and since the cause of a whole is primarily the cause of the components 

making up the whole, each of the possibly existent components would be part of 
ji the cause of the existence of itself. 139